Genetic diversity and conservation biology of peregrina populations in Sinai,

Inauguraldissertation zur

Erlangung des akademischen Grades eines

Doktors der Naturwissenschaften (Dr.

rer. nat.) der

Mathematisch-Naturwissenschaftlichen

Fakultät der

Ernst-Moritz-Arndt-Universität Greifswald

vorgelegt von

Mohamed A. EL Dadamouny A. Sheteiwy

geboren am 03.05.1981

in El Manzala, Ägypten

Greifswald, März 2018

Dekan: Prof. Dr. Werner Weitschies

1. Gutachter : Prof. Dr. Christoph Oberprieler Universität Regensburg, Germany

2. Gutachter: Prof. Dr. Martin Schnittler Universität Greifswald, Germany

Tag der Promotion: 21.08.2018.

Dedication I dedicate my dissertation to my family. To my father, my mother, and to my wife Mrs. Riham Sheteiwy My daughters Sama, Hana, and Jana, and to the spirit of my grandmother

Preface

Since my undergraduate stage (1998–2002) at Faculty of Science in Ismailia – Egypt, I used to visit for field work. I was attracted to the highly diverse flora of Sinai, which includes around 2000 species. Most of these species are medicinal that used in the folk medicine and considered as one of the income resources to the in this interesting peninsula. One of these medicinal plants was Moringa peregrina, which was standing with Acacia species and Crataegus x Sinaica for several years facing the impacts of high aridity of the area and over collection. These threats led to a gap in the population dynamics. One of the common observations at that time was the poor recruitment due to the low establishment. I did my graduation project (2001–2002) on that point trying to increase the establishment rate with a decreasing of the mortality rate for the germinated seedlings. From 2002 to 2004, I worked within a project aimed to conserve the remaining populations. I got a great chance to understand more about the tree biology (e.g. phenological stages starting from budding and flowering time, fruiting, and etc.). From 2004–2009, I did also my M.Sc. project on the population dynamics of this valuable species. For that reason, its declining populations are located under threats of climate change and anthropogenic impacts. The response of the tree towards such external threats was unclear with a continuous missing in the recruitment and then a continuous decline in its population size. Establishing the conservation programs could help in mitigating the adverse impacts of these threats. But, the tree itself should have the capability to adapt and to overcome towards these external problems if the current populations are genetically diverse. Therefore, we aimed in my Ph.D. project to find a solution for the gap on its reproductive biology and genetic diversity within and among its populations that needed for conservation programs. As an outputs for all works done over the past years under the supervision of Prof. Martin Schnittler, four scientific papers (two accepted papers and two manuscripts) and one review are achieved and included in this cumulative thesis.

List of publications included in this cumulative dissertation Peer-Reviewed Publications Dadamouny MA and Schnittler M (2016) Trends of climate with a rapid change in Sinai, Egypt. Journal of water and climate change 7(2):393–414. IWA doi: 10.2166/wcc.2015.215 Dadamouny MA, Unterseher M, König P, and Schnittler M (2016) Population performance of Moringa peregrina (Forssk.) Fiori (Moringaceae) at Sinai Peninsula, Egypt in the last decades: consequences for its conservation. Journal for Nature Conservation 34:65–74 Manuscripts Dadamouny MA and Schnittler M. Demographic and size structure status of the threatened populations of Moringa peregrina trees at Sinai Peninsula, Egypt.

Mohamed Dadamouny, Anja Klahr, Franziska Eichhorn, and Martin Schnittler. Relictic populations of the endangered tree Moringa peregrina (Moringaceae) in Sinai, Egypt are still genetically diverse. 4

Genetic diversity and conservation biology of Moringa peregrina populations in Sinai, Egypt

Contents

Contents Page

Title page ------1 Defense committee ------2 Dedication ------3 Preface ------4 Table of Contents ------5 Chapter 1: Summary ------7 1.1. Englisch Summary ------7 1.2. German Summary ------9 Chapter 2: Introduction ------12 Background ------12 1. Species identification ------12 1.1. Taxonomic position of M. peregrina ------12 1.2. Distribution of M. peregrina ------12 1.3. Morphology of M. peregrina ------13 1.4. Anatomy of M. peregrina ------17 1.5. Phytochemical analyses of M. peregrina ------18 1.6. Seed viability and of M. peregrina ------18 1.7. Uses of M. peregrina ------19 2. Problem identification and looking for solutions ------19 3. Goals of the thesis ------21 References ------22

Chapter 3: Publications ------24 3.1. Trends of climate with a rapid change in Sinai, Egypt ------25 Abstract ------25 Introduction ------25 Study area ------27 Material and methods ------29 Results ------30 Discussion ------36 Conclusion ------41 References ------42 Appendix I ------47 Appendix II ------50 3.2. Population performance of Moringa peregrina (Forssk.) Fiori (Moringaceae) at Sinai Peninsula, Egypt ------52 Abstract ------52 Introduction ------52 Material and methods ------54 Study area ------54 Results ------56 Discussion ------63 5

Contents Genetic diversity and conservation biology of Moringa peregrina

Conclusion ------65 References ------66 Appendix III ------69 Appendix IV ------71 3.3. Demographic and size structure status of the threatened populations of Moringa peregrina trees at Sinai Peninsula, Egypt ------78 Abstract ------78 Introduction ------78 Material and methods ------79 Study area ------79 Results ------82 Discussion ------86 Conclusion ------88 References ------89 Appendix V ------92 Appendix VI ------94 3.4. Relictic populations of the endangered tree Moringa peregrina (Moringaceae) in Sinai, Egypt are still genetically diverse ------96 Abstract ------96 Introduction ------96 Material and methods ------97 Study area and sampling ------97 Results ------99 Discussion ------103 References ------106 Appendix VII ------109 Appendix VIII ------113

Chapter 4: Conclusions and perspectives ------117 Conclusions ------117 Declaration ------120 Curriculum Vitae ------125 Acknowledgements ------130

6

Genetic diversity and conservation biology of Moringa peregrina 1 populations in Sinai, Egypt Genetische Diversität und Naturschutzbiologie der Populationen von Moringa peregrina in Sinai, Ägypten. Summary

2.1. English Summary (Englische Zusammenfassung) The populations of the tree Moringa peregrina growing at Sinai Peninsula (Egypt) are threatened by climate change and anthropogenic impacts. The response of the tree towards such external threats was unclear, but missing recruitment and a continuous decline in its population size indicate that the populations are in danger. Establishing conservation programs could help in mitigating the adverse impacts of these threats. But, the tree itself may have the capability to adapt if the remaining populations are genetically diverse. Therefore, the goal of this Ph.D. study was to gain insights into the reproductive biology and genetic diversity within and among the populations of M. peregrina.

First, we studied the threat situation for the remaining M. peregrina populations based on data collected through fieldworks in 2014 and 2015, which resulted in three publications. The first, dealing with trends of the climate of the study area (Chapter 3.1), was published in Journal of Water and Climate Change. It proves that the climate in Sinai changed during the last 45 years towards increasing aridity due to the increase of the mean annual temperature by +2.3OC and the decline precipitation, with rainfalls becoming rarer, less predictable and fluctuate more in extent. Flash floods after several years of drought became very common, which could affect the dynamics of the remaining populations of M. peregrina. The second publication (Chapter 3.2) focuses on the population performance and consequences for conservation of the species, published in Journal for Nature Conservation. Based on these results the IUCN status for M. peregrina was updated to be an endangered species on the IUCN Red list of the threatened species. In the third publication (Chapter 3.3, unpublished manuscript) the demography and recruitment status of M. peregrina populations was studied, which helped to evaluate the change in these populations over time. Adapted to arid , our results proved that Moringa peregrina is one of the few subtropical trees displaying annual growth rings that allow age determination. The relationship between age and tree radius could be used to determine the age structure of four populations, which showed prolonged periods of missing recruitment.

Second, at the lab of the Dept. General and Plant Systematics and at the in Greifswald, all experiments for genetic assessments are done. Because Moringa is rich in secondary metabolites and phenolics, we faced a challenge in extracting a pure DNA required for AFLP (the first proposed genotyping method). Later, different DNA isolation methods were tested to overcome the obstacles caused by phenolics and sugars, an AFLP protocol that worked well with the cultivated seedlings at the botanical garden in

7

Summary Genetic diversity and conservation biology of Moringa peregrina

Greifswald was established, markers for the Internal Transcribed Spacer (ITS) were tested, and finally, SSR (microsatellite) markers were established. To optimize DNA extraction, the method of Doyle and Doyle was modified and optimized. This is an ideal method for obtaining a non-fragmented DNA that could be used for AFLP. In addition, two other DNA extraction methods; (KingFisher Flex robot using Omega M1130 extraction Kits, and spin columns and 96-plates using Stratec kits). Although we achieved similar results for both Robot kits (Omega) and Stratec kits, the amplification for most of the samples extracted with Robot did not work, therefore the Stratec kit was the method of choice as it has also a lower cost, combined with a high quality of DNA. For ITS, no polymorphism was found for 28 samples of M. peregrina from Sinai (sequences submitted to GenBank). However, since microsatellite markers of M. peregrina were not known, it was a challenge to try a cross amplification from other species with well-known microsatellite primers. Cross-amplification of 16 primers known from the related species M. oleifera was tested, and three multiplex PCR reactions were established after testing different annealing temperatures and different primers concentrations. This included 13 primers out of the 16 investigated markers which gave a reliable band. All methods used for genetic assessments for the different Moringa species are compiled in a comparative review to look for connections between the different Moringa species. For Moringaceae, M. oleifera and M. peregrina are closely related to each other. Both species have slender trunks, with thick, tough bark and tough roots and bilaterally symmetrical flowers with a short hypanthium. Both species have also a pubescent and small pollen grains, their distribution ranges overlap. However, the seeds in M. peregrina are unwinged and the leaf axes persist over the dry season, whereas seeds in M. oleifera are usually winged and leaf axes are .

All genetic data were presented in the fourth publication (Chapter 3.4 as a manuscript). All but one SSR markers used in this study are highly informative However, the degree of polymorphy varied considerably within the 13 markers used. The Probability of Identity (PI) for all loci was 2.6 x 10-9 with high resolution. The percentage of polymorphic loci for all populations was 88.5±2.2; figures for single populations were 92.3%, 84.6%, 84.6%, and 92.3% for the wadis WM, WA, WF, and WZ, respectively. The genotype accumulation curve as well demonstrated that 7–8 markers were necessary to discriminate between 100% of the multilocus genotypes. Significant departures from HWE were detected for eight loci (P < 0.001), probably due a high degree of inbreeding within population. The observed (HO) and expected (HE) heterozygosities ranged from 0 to 0.86 and from 0 to 0.81, respectively. However, for the pooled population, excluding the monomorphic locus MO41, HO and HE ranged from 0.069 to 0.742 and from 0.126 to 0.73 with averages of 0.423 and 0.469,

respectively. The mean of FST was 0.133, indicating that, due to the long generation time of M. peregrina, there is still relatively little differentiation between the four remaining populations. An analysis of molecular variance (AMOVA) revealed that the old populations of M. peregrina are still genetically diverse where 75% of variance was recorded within individuals and 83% within populations. An analysis with STRUCTURE, varying the parameter K between 1 and 7, revealed the most pronounced genetic structure for K=3, thus uniting the populations from two neighboured wadis (W. Agala and W. Feiran). The three groups seem to be now genetically isolated. (They may be remainders of a formerly contiguous population, especially when considering the change towards a drier climate in Northern within the last 6000 years). Six clones of each two individuals collected from the same wadi were found, pointing to vegetative dispersal via broken twigs, which 8

Genetic diversity and conservation biology of Moringa peregrina Summary may have rooted after flash floods. It may be an alternative mode of reproduction under harsh conditions. Our data reveal a low gene flow between three of the four wadis, suggesting that the trees are relictual populations. In general, conservation of populations from the three genetically most diverse wadis and cross-breeding of trees within a reforestation program is recommended as an effective strategy to ensure the survival of M. peregrina at Sinai, Egypt.

2.2. German Summary

Die noch erhaltenen Populationen von Moringa peregrina, einem an Trockenheit angepassten laubabwerfenden Baum, der auf der ägyptischen Sinai-Halbinsel vorkommt, sind durch Klimawandel und anthropogene Einflüsse bedroht. Die Reaktion des Baumes auf solche externen Bedrohungen war unklar, aber fehlende Rekrutierung und ein kontinuierlicher Rückgang der Populationsgröße deuten darauf hin, dass die Populationen in Gefahr sind. Die Einrichtung von Schutzprogrammen könnte dazu beitragen, die negativen Auswirkungen dieser Bedrohungen zu mildern. Eventuell kann sich die Art an geänderte Umweltbedingungen anpassen, wenn die restlichen Populationen genetisch verschieden sind. Daher ist das Ziel dieser Promotionsarbeit, Einsichten in die Reproduktionsbiologie und genetische Vielfalt innerhalb und zwischen den Populationen von M. peregrina gewinnen.

Zunächst haben wir die Bedrohungslage für die restlichen Populationen von M. peregrina auf der Grundlage von Daten untersucht, die 2014 und 2015 in Feldforschungen erhoben wurden. Daraus ergaben sich drei Publikationen. Die erste, die sich mit den Trends des Klimas des Untersuchungsgebiets befasst (Kapitel 3.1), wurde im Journal of Water and Climate Change veröffentlicht. Es zeigt, dass sich das Klima im Sinai in den letzten 45 Jahren in Richtung zunehmender Trockenheit durch den Anstieg der mittleren Jahrestemperatur um +2.3° C und den Rückgang der Niederschläge verändert hat, wobei Regenfälle seltener, weniger vorhersehbar und schwankender werden. Sturzfluten nach mehreren Jahren Trockenheit wurden sehr häufig, was die Dynamik der verbleibenden Populationen von M. peregrina beeinträchtigen könnte. Die zweite Veröffentlichung (Kapitel 3.2, Journal for Nature Conservation) konzentriert sich auf die Populationsleistung und die Konsequenzen für den Artenschutz. Basierend auf diesen Ergebnissen wurde der IUCN-Status für M. peregrina als gefährdete Art auf der Roten Liste der IUCN der bedrohten Arten aktualisiert. In der dritten Veröffentlichung (Kapitel 3.3, unveröffentlichtes Manuskript) wurde der Demographie- und Rekrutierungsstatus von M. peregrina-Populationen untersucht, was dazu beitrug, die Veränderung dieser Populationen im Laufe der Zeit zu bewerten. Unsere Ergebnisse zeigen, dass Moringa peregrina einer der wenigen subtropischen Bäume ist, die jährliche Wachstumsringe aufweisen, womit eine Altersbestimmung möglich wird. Die Beziehung zwischen Alter und Baumradius könnte verwendet werden, um die Altersstruktur von vier Populationen zu bestimmen, die längere Perioden fehlender Rekrutierung zeigten.

Zweitens werden im Labor der Arbeitsgruppe Allgemeine & Spezielle Botanik und im Botanischen Garten in Greifswald Experimente für die Bestimmung der genetischen Diversität durchgeführt. Da Moringa reich an Sekundärmetaboliten und Phenolen ist, war es eine Herausforderung, reine DNA extrahieren, die für AFLP (die

9

Summary Genetic diversity and conservation biology of Moringa peregrina erste in Betracht gezogene Genotypisierungsmethode) benötigt wird. Später wurden verschiedene DNA- Isolierungsmethoden getestet, um die durch die hohe Konzentration von Phenolen und Zucker verursachten Hindernisse zu überwinden. Es wurde ein AFLP-Protokoll entwickelt, das mit den kultivierten Keimlingen im Botanischen Garten in Greifswald gut funktionierte; Marker für den Internal Transcribed Spacer (ITS) wurden getestet; und schließlich wurden SSR (Mikrosatelliten) Marker etabliert. Um die DNA-Extraktion zu optimieren, wurde die Methode von Doyle und Doyle modifiziert und optimiert. Dies ist eine ideale Methode, um eine nicht fragmentierte DNA zu erhalten, die für AFLP verwendet werden könnte. Darüber hinaus zwei weitere DNA-Extraktionsmethoden getestet (KingFisher Flex Roboter mit Omega M1130 Extraktions-Kits und Spin-Säulen und 96-Platten mit Stratec-Kits). Obwohl wir ähnliche Ergebnisse sowohl für die Robot Kits (Omega) als auch für die Stratec Kits erzielten, funktionierte die Amplifikation für die meisten mit Robot extrahierten Proben nicht, daher war das Stratec Kit die Methode der Wahl, da es geringere Kosten mit einer hohen DNA-Qualität kombiniert. Für ITS wurde für 28 Proben von M. peregrina aus Sinai (Sequenzen wurden an GenBank übermittelt) kein Polymorphismus gefunden. Da Mikrosatelliten-Marker von M. peregrina nicht bekannt sind, war es eine Herausforderung, eine Kreuz-Amplifikation von anderen Spezies mit bekannten Mikrosatelliten-Primern zu versuchen. Eine Kreuzamplifikation von 16 Primern, die aus der verwandten Art M. oleifera bekannt sind, wurde getestet, und drei Multiplex-PCR-Reaktionen wurden nach dem Testen verschiedener Anlagerungstemperaturen (annealing) und unterschiedlicher Primerkonzentrationen etabliert. Die Reaktionen funktionierten mit 13 von 16 untersuchten Markern, die eine zuverlässige Bande ergaben. Alle Methoden zur genetischen Bewertung der verschiedenen Moringa-Arten werden aus der Literatur zusammengestellt und in einer vergleichenden Übersicht zusammengefasst (Kapitel 3.4, unveröffentlichtes Manuskript), um nach Gemeinsamkeiten zwischen den verschiedenen Moringa-Arten zu suchen. innerhalb der Familie Moringaceae sind M. oleifera und M. peregrina eng miteinander verwandt. Beide Arten haben schlanke Stämme mit dicker, zäher Rinde und harten Wurzeln und bilateral-symmetrische Blüten mit kurzem Hypanthium. Beide Arten haben auch einen behaarten Fruchtknoten und relativ kleine Pollenkörner, und ihre Verbreitungsbereiche überlappen sich. Allerdings sind die Samen in M. peregrina ungeflügelt und die Blattachsen (Rachis) bleiben während der Trockenzeit am Baum, während die Samen in M. oleifera gewöhnlich geflügelt sind und die Blattachsen abfallen.

Alle genetischen Daten wurden in der vierten Publikation (Kapitel 3.4 als Manuskript) präsentiert. 12 von 13 SSR-Marker, die in dieser Studie verwendet wurden, sind sehr informativ. Der Grad der Polymorphie variiert innerhalb der 13 verwendeten Marker für verschiedene Genorte) beträchtlich. Die Wahrscheinlichkeit der Identität (PI) für alle Loci betrug 2,6 × 10-9. Der Prozentsatz der polymorphen Loci für alle Populationen betrug 88,5 ± 2,2; Die Zahlen für einzelne Populationen betrugen 92,3%, 84,6%, 84,6% und 92,3% für die Wadis WM, WA, WF bzw. WZ. Eine Akkumulationskurve der Genotypen zeigte, auch, dass 7-8 Marker notwendig waren, um zwischen 100% der Multilocus-Genotypen zu unterscheiden. Signifikante Abweichungen von HWE wurden für acht Loci gefunden (P <0,001), wahrscheinlich aufgrund eines hohen Inzuchtgrades innerhalb der Population. Für jedes der vier untersuchten Wadis lag die beobachtete (HO) und erwartete (HE) Heterozygosität zwischen 0 und 0,86 bzw. zwischen 0 und 0,81. Für die gepoolte Population mit Ausnahme des monomorphen

10

Genetic diversity and conservation biology of Moringa peregrina Summary

Locus MO41 lagen HO und HE jedoch zwischen 0,069 und 0,742 und zwischen 0,126 und 0,73 mit

Durchschnittswerten von 0,423 bzw. 0,469. Der Mittelwert der genetischen Differenzierung (FST) betrug 0,133, was darauf hindeutet, dass aufgrund der langen Generationszeit von M. peregrina immer noch relativ wenig zwischen den vier verbleibenden Populationen unterschieden wird. Eine Analyse der molekularen Varianz (AMOVA) ergab, dass die alten Populationen von M. peregrina immer noch genetisch unterschiedlich sind, wobei 75% der Varianz innerhalb von Individuen und 83% innerhalb von Populationen aufgezeichnet wurden. Eine Analyse mit dem Programm STRUCTURE variierte den Parameter K zwischen 1 und 7 und zeigte die stärkste genetische Struktur für K = 3, wodurch die Populationen aus zwei benachbarten Wadis (W. Agala und W. Feiran) vereinigt wurden. Die drei Gruppen scheinen nun genetisch isoliert zu sein. (Sie können Reste einer ehemals zusammenhängenden Population sein, besonders wenn man die Veränderung in Richtung eines trockeneren Klimas in Nordafrika in den letzten 6000 Jahren betrachtet). Sechs Klone von jeweils zwei Individuen, die aus demselben Wadi gefunden wurden, die auf vegetative Verbreitung durch gebrochene Zweige hinwies, die nach Sturzfluten verwurzelt sein können. Unsere Daten zeigen einen geringen Genfluss zwischen drei der vier Wadis, was darauf hindeutet, dass die Bäume reliktische Populationen sind. Zusammenfassend wird die Erhaltung der Populationen aus den drei genetisch sehr unterschiedlichen Wadis und die Kreuzung von Bäumen innerhalb eines Wiederaufforstungsprogramms als wirksame Strategie empfohlen, um das Überleben von M. peregrina im Sinai, Ägypten, zu gewährleisten.

11

2 Genetic diversity and conservation biology of Moringa

peregrina populations in Sinai, Egypt Introduction

1. Species identification 1.1. Taxonomic position of M. peregrina Moringa, the only genus within the family Moringaceae (the order ) occurs with 13 species throughout the dry tropics (Al Kahtani and Abou-Arab, 1993; Quattrochi, 2000; Olson and Carlquist, 2001). According to Boulos (1999) and from the system by Melchior (1964), adapted than system of Engler, M. peregrina is classified as follows: Kingdom: Plantae Phylum: Spermatophyta Subdivision: Angiospermae, Dicotyledoneae Class: Rosopsida Order: Brassicales Family: Moringaceae Genus: Moringa Species: M. peregrina (Forssk.) Fiori

1.2. Distribution of M. peregrina M. peregrina is native to the area around the and the with its gulfs. The tree is recorded from NE Africa (Egypt, , , , Eritrea, and ) and SW Asia (Palestine, Israel, Jordan, W and SW Saudi Arabia, Yemen, Oman, UAE). In Egypt: South Sinai and Red Sea Mountains (Gebel Shindodai, Samiuki, Nugrus, and Shayeb). Figure 1 shows the geographical distribution of M. peregrina worldwide, which is focused around the Red Sea.

Fig. 1. Distribution map of M. peregrina around the Red Sea, and Arabian Peninsula (Dadamouny et al. 2016). 12

Introduction Genetic diversity and conservation biology of Moringa peregrina

According to Kassas and Zahran (1962) and (1971), M. peregrina is confined to the feet of the mountains that are higher than 1300-1500m.; extending from Gebel Abou-Dukhan (lat. 27º 20′ N) to Gebel El-Faryid (lat. 23º 30′ N) (El Hadidi et al., 1991). M. peregrina prefers the upstream parts of wadis draining the slopes of the higher mountains (Zahran and Willis, 2009). Besides, M. peregrina, here often growing shrubby, occurs in patches that cover limited areas of the upstream gulches of the drainage systems. These gulches collect water at the foot of the higher mountains (Dadamouny, 2009). The ground where M. peregrina grows is usually covered with coarse rock debris or detritus, as typical for gulches and ravines at the mountain bases and slopes. The species is known as well in South Sinai from a limited area at Feiran Oasis Mountains, and in a few wadis (W. Agala, W. Meir and W. Zaghra). Also, few, now vanished or strongly declining populations were distributed in the Eastern Desert and Gebel Elba in south-eastern Egypt (Abd El-Wahab, 1995; Dadamouny, 2009; et al., 2016). Moreover, the west-facing escarpments of Gebel Serbal were rich in M. peregrina trees (Danin, 1999) that but these populations ceased due to extreme droughts (Dadamouny et al., 2016).

1.3. Morphology of M. peregrina M. peregrina (Forssk. Fori) is a medium sized deciduous tree (Fig. 2) 3 to 15 m tall (Boulos, 1999; El- Hadidi et al., 1991; Dadamouny. 2009). Its main root is thick (Fig. 3a and 3b) and the stem is forked or multistemmed (Fig. 2). Its branches and stem are brittle with corky bark. Therefore, trunk is erect, terete, branched, and branches is divaricated or ascending, slender, forming avoid or abavoid crown, green-glaucaus (El-Hadidi et al., 1991). The leaves of M. peregrina (Fig. 3c) are feathery, pale green, compound tri-pinnate, 30-60 cm long with many small leaflets which are 1.3-2.5 cm and 0.3-0.6 cm wide (Boulos, 1999). The leaflets are early deciduous, simple, petiolate, glabrous on both surfaces, the blade is ovate- oblanceolate, margin entire, and apex obtuse, sometimes mucronate (El-Hadidi et al., 1991; Dadamouny, 2009). When M. peregrina seedlings start out, they have broad leaflets and a large tuber. Through many dry seasons, the shoot dies back below ground to the tuber. As the plant gets older, the leaves get longer and longer, but the leaflets get smaller and smaller and more widely spaced. Adult trees produce leaves with a full complement of tiny leaflets, only to drop them as the leaf matures (Olson, 1999).

Flowers of M. peregrina are fragrant, white, creamy, pinkish to pale, 2.5 cm in diameter, born in sprays, with five yellow protruding the flower (Fig. 4a) Also, flowers are pedicellate and arranged in panicles (see floral diagram in Fig. 4b). The are oblong-lanceolate, acuminate, and whitish. The colour of the is white-pinkish to pale (El-Hadidi et al., 1991; Dadamouny, 2009). The fruits (Figures 4c and 4d) or pods of M. peregrina are pendulous ridged, brown, triangular. The pod (Figure 2.c) splits lengthwise into 3 parts containing about 20-25 trigonous seeds embedded in the pith. Pods are tapering at both ends, 9-ribbed. The seeds of M. peregrina are dark brown with 3 papery wings (Figure 4.e). Zahran and Willis (2009) stated that the pendulous pods ripen in October. The angled nut-like white seed [behen nut] are bitter-sweet, with a nauseous taste and rich in oil (ben-oil) (Täckholm, 1974; Dadamouny et al., 2012).

13

Genetic diversity and conservation biology of Moringa peregrina Introduction

Fig. 2. Morphology of M. peregrina showing the different pattern of growth for the tree stem (forked; multi-forked, or multistemmed tree, candelabra tree or ).

14

Introduction Genetic diversity and conservation biology of Moringa peregrina

Fig. 3. The thick root of M. peregrina in W. Agala (a) under the impact of soil erosion after several years of droughts and in W. Zaghra (b) under the impact of drought and infection. The leaves for one the seedlings at the botanical garden in Greifswald, Germany. The leaves are compound tri-pinnate as shown also for new seedling recorded in W. Zaghra in 2014 and disappeared in 2015 (Dadamouny et al., 2016).

The pedicellate flowers (Fig. 4a) are arranged in panicles (Fig. 4b shows its floral diagram) and produce fruits or pods (Fig. 4c) from May up to July; each pod contains 5–15 seeds (Fig. 4d, see El-Hadidi et al., 1991). Mature trees produce about 150 pods, each containing 10 to 15 seeds (1700 seeds/kg pods). December and January are the dry period, where M. peregrina sheds its leaflets. It is a highly nutritious plant that produces economically valuable oil making up for to 54% of dry pods (Somali et al., 1984), besides its other benefits as a medicinal plant, water purifier, and animal fodder (Abd El-Wahab, 1995). The roots, leaves, flowers and seeds of M. peregrina are used as medication for tumors (Hartwell, 1971). Pods act as a de-wormer and treat liver and spleen problems and pains of the joints. Due to the high content of proteins, , vitamins (A and C), minerals (Ca, P, Fe and Cu), and fibres of pods, it can treat malnutrition and (Dadamouny, 2009). The leaves of M. peregrina are also a source of vitamins and proteins. It is used as well against headaches and to stop bleeding. There is an anti-bacterial and anti-inflammatory effect when it applied to wounds or insect bites (Price, 1985; Freiberger et al., 1998; Dadamouny et al., 2012).

15

Genetic diversity and conservation biology of Moringa peregrina Introduction

(b) (a)

(d) (c)

c

(e) (f)

Fig. 4. Flowers, pods, and seeds of M. peregrina, Flowers of a tree from Wadi Meir, South Sinai (a), floral diagram (b), fruits or pods started in ripening (c), early pods (d) opened and late flowers which usually produce pods (e) and seeds, these are easy to germinate rate in moist conditions without any pre-treatment (f).

16

Introduction Genetic diversity and conservation biology of Moringa peregrina

The monogeneric family (Moringaceae) occupies a systematic position in or near the Capparales, because of the shared presence of glucosinolates and mustard-oils, along with several other characters, such as zygomorphic flowers, the presence of a gynophore, parietal placentation, and lack of endosperm (Pax, 1891, 1936; Dahlgren, 1980, 1989; Cronquist, 1981; Takhtajan, 1997; Ronse Decraene et al., 1998). Moringa appears as part of an extended Capparales. Rodman et al. (1993) recognized a core group of Capparales (including Capparaceae, Tovariaceae, Brassicaceae and Resedaceae) and only two independent origins of mustard-oil producing plants were postulated, one leading to the Capparales and the other to Drypetes (Euphorbiaceae). 1.4. Anatomy of M. peregrina Based on the gross appearance, the genus of Moringa is divided into four classes; bottle trees, slender trees, sarcorhizal trees and tuberous trees. In case of slender trees (e.g. M. peregrina), the stem is characterized by a preponderance of libriform fibres that show little seasonal variation in shape with little axial parenchyma. Also in favourable season, the early wood libriform fibres are sometimes replaced by confluent aliform paratracheal parenchyma (Carlquist, 1988). M. peregrina is a slender trees showing alternating bands of libriform fibres and paratracheal axial parenchyma. However, the parenchyma bands are never wider than adjacent bands of fibres. This predominance of libriform fibres makes the roots one of the toughest in the genus. The fibres of M. peregrina may occur in rings separated by aliform to the confluent aliform parenchyma. Rays are differing through life forms mainly in size and proportion of upright to square to procumbent cells. The shortest multiseriate rays were found (Olson and Carlquist 2001). Tyloses were also observed in all Moringa species. Storing can be observed in all Moringa species but is most apparent in area of extensive axial parenchyma or short libriform fibres (Olson and Carlquist 2001).

At the first hand, the stem of M. peregrina in the cross-section is terete (Al-Gohary and Hajar, 1996). Epidermal cells tangentially to radially elongated, cuticle thick. Trichomes eglandular and unicellular. Cortex with 2-5 layers of small parenchyma cells containing cluster crystals and starch grains followed by 5-8 layers of large thin-walled parenchyma cells; Vascular cylinder crowned with isolated patches of protophloem which are stratified into hard fibre patches. Vascular tissue in the form of dicttyostele which is composed of slightly separated vascular bundles. Vascular cambium with 3-4 layers of radially arranged rectangular cells. Xylem vessels with lignified reticulate, spiral and pitted thickenings. Pith formed of thin walled parenchymatous cells containing cluster crystals and starch grains. On the other hand, the epidermal cells in the leaf of M. peregrina are tangentially and radially elongated. Trichomes eglandular and unicellular. Mesophyll of dorsoventrally type. Palisade tissue of 1-2 layers which are discontinuous adaxially at the midrib region. Mid veins are crescent-shaped and surrounded by parenchymatous sheath. Mechanical reinforcements (collenchyma) were recorded abaxial and adaxial, at the midrib region.

It is obvious that various anatomical changes occur in stems and leaves of M. peregrina in response to variations of soil moisture content (Al-Gohary and Hajar, 1996). The diameter of stem increased gradually as water supply increased. The number of parenchymatous layers of cortex progressively also increased with increase in the available moisture. The radial section of M. peregrina wood (SEM) shows vessels and fibres cells that characterized by the scalable border pitting system that increases its wood permeability. In addition, ray parenchyma can be observed (Hindi, 2017).

17

Genetic diversity and conservation biology of Moringa peregrina Introduction

1.5. Phytochemical analyses of M. peregrina Table 1. Nutritional Analysis of M. peregrina pods, fresh raw leaves, and dried The kernel (seed) of M. peregrina contains leaf powder per 100 grams of edible portion (Price, 1985; Freiberger et al., 1998; Dadamouny, 2009). 1.8% moisture, 54.3% oil, 22.1% protein, 3.6% Fresh Dried Leaf Components Pods Powder fibre, 15.3% and 2.5% ash. Leaves Moisture (%) 86.90% 75% 7.50% Calories 26.0 92.0 205.0 Moreover, the composition and characteristic of Protein (g) 2.5 6.7 27.1 Fat (g) 0.1 1.7 2.3 the extracted oil were determined (Somali et al., Carbohydrate (g) 3.7 13.4 38.2 Fibre (g) 4.8 0.9 19.2 1984). Gas liquid chromatography of methyl Minerals (g) 2.0 2.3 - (mg) 30.0 440.0 2003.0 esters of the fatty acids showed the presence of Magnesium (mg) 24.0 24.0 368.0 Phosphorous (mg) 110.0 70.0 204.0 14.7% saturated fatty acids and 84.7% Potassium (mg) 259.0 259.0 1324.0 Copper (mg) 3.1 1.1 0.6 unsaturated fatty acids. The fatty acid (mg) 5.3 0.7 28.2 Oxalic acid (mg) 10.0 101.0 0.0 composition explains as follows: Palmitic Sulphur 137.0 137.0 870.0 VITAMINS CONTENTS (9.3%), Palmitoleic (2.4%), Stearic (3.5%), - B carotene (mg) 0.1 6.8 16.3 Vitamin B – Choline (mg) 423.0 423.0 Oleic (78.0%), Linoleic (0.6%), Araachidic Vitamin B1 – Thiamine (mg) 0.1 0.2 2.6 Vitamin B2 – Riboflavin (mg) 0.1 0.1 20.5 (1.8%) and Behenic (2.6%). Moreover, M. Vitamin B3 – Nicotinic Acid (mg) 0.2 0.8 8.2 – Ascorbic Acid (mg) 120.0 220.0 17.3 peregrina is the uppermost for antioxidant Vitamin E –Tocopherols Acetate (mg) - - 113 AMINO ACIDs CONTENTS (Yang et al., 2006a and b). The leaves and green Arginine (mg) 360.0 406.6 1325 Histidine (mg) 110.0 149.8 613 branches of M. peregrina are rich in minerals, Lysine (mg) 150.0 342.4 1325 Tryptophan (mg) 80.0 107.0 425 amino acids, vitamins (Table 1), carbohydrates, Phenylalanine (mg) 430.0 310.3 1388 Methionine (mg) 140.0 117.7 350 and phenolic compounds (El-Lamey, 2015). Threonine (mg) 390.0 117.7 1188 Leucine (mg) 650.0 492.2 1950 For that reason, Moringa is the tree of life. Isoleucine (mg) 440.0 299.6 825 Valine (mg) 540.0 374.5 1063

Table 2. Nutritional value of M. peregrina, compared to concentrations in common grocery products

FRESH LEAVES DRIED LEAF POWDER 4 times Vitamin A of Carrots 10 times Vitamin A of Carrots 7 times Vitamin C of Oranges ½ times Vitamin C of Oranges 4 times Calcium of Milk 17 times Calcium of Milk 3 times Potassium of Bananas 15 times Potassium of Bananas ¾ times Iron of Spinach 25 times Iron of Spinach 2 times Protein of Yogurt 9 times Protein of Yogurt

1.6. Seed viability and germination of M. peregrina Soaking of M. peregrina seeds in diluted tetrazolium salt (0.007 gm/L), an experiment to check for vital cell membranes, showed that all seeds are viable (Abd El-Wahab, 1995; Dadamouny, 2009). Although M. peregrina has hard woody testa, the highest percentage (92%) of germination was obtained when M. peregrina seeds were pre-soaked in water up to 24 hours (Dadamouny, et al., 2012). The seeds of M. peregrina (Fig, 4f) germinated within a few days without any pre-treatment. After ten days at 25ºC, 90 % of M. peregrina seeds were germinated. Thus, propagation of M. peregrina via seeds in addition to cuttings was recommended (Abd El-Wahab, 1995; Dadamouny, 2009).

18

Introduction Genetic diversity and conservation biology of Moringa peregrina

1.7. Uses of M. peregrina M. peregrina has different uses in medicine, , and as a plant growth hormone (see Dadamouny, 2009). People of the Arab Peninsula and Sinai use different parts, mainly seeds, roots and leaves of M. peregrina for treatment of tumors, inflammations, malnutrition, and sexual dysfunction. Pods act as a de- wormer and treat liver and spleen problems and pains of the joints. Due to the high protein and fibre content of pods, it playing a role treating malnutrition and diarrhea. The root is used also for dropsy. Moreover, roots are bitter as a tonic to the body and lungs, expectorant, mild diuretic and stimulant in paralytic afflictions, epilepsy and hysteria. The leaves of M. peregrina are applied as poultice to sores, for headaches, and said to have purgative properties and stop bleeding. There is an anti-bacterial and anti-inflammatory effect when applied to wounds or insect bites. Extracts can be used against bacterial or fungal skin complaints. Leaf tea treats gastric ulcers and diarrhea. Bark, leaves and roots are acrid, pungent, and are taken to promote digestion (Freiberger et al., 1998). The flower juice of M. peregrina is useful for urinary problems. Moreover, it improves the quality and flow of mothers’ milk.

The seed oil of M. peregrina is used for diarrhea and conversely it has a laxative effect. It deserves to be an important part in our diet as a concentrated source of food energy and nutrients. The seeds of M. peregrina were the source of the oil used by the since the time of old and middle dynasties (300-200 B.C). The refined oil obtained from Moringa seeds has a yellowish colour, a sweet taste and odourless, for this reason it was much estimated for preparing cosmetics (Table 3). The bright yellow oil with a pleasant taste has been compared in quality with oil. The kernel contains 35-40% by weight of oil (Folkard & Sutherland, 2005). Oil content of M. peregrina seeds from Saudi Arabia was 49.8%, was a refractive index (40 °C) 1.460, density (24 °C) 0.906, acidity (as oleic) 0.30%, iodine value 69.6, saponification number 185 and peroxide value 0.4 meq/kg. Moringa peregrina seed oil was found to contain high levels of oleic (70.52%), followed by gadoleic (1.5%), while the dominant saturated acids were palmitic (8.9%) and stearic (3.82%). α- γ- and δ-tocopherols were detected at levels of 145,58 and 66 mg/kg respectively (Tsaknis, 1998).

Table 3. Chemical properties of M. peregrina oil (Somali et al., 1984) and M. oleifera oil (Ashfaq et al., 2012). Moringa peregrina Oil Name Fiori Oil Ben Oil Refractive index 1.4610 1.4571 Iodine Value 69.5 68.63 Saponification Value mg KOH/g 182.9 181.4 Acid Value 0.04 0.81 Unsaponifiable Matter % 0.3 0.74 Saturated Acids % 14.7 12.4 Myristic (C14:0) Trace -- Palmitic (C16:0) 9.3 2.97 Stearic (C18:0) 3.5 3.97 Behenic (C22:0) 2.6 4.86 Lignoceric (C24:0) -- 0.6 Unsaturated Acids % 84.7 82.44 Oleic (C18:1) 78.0 81.73 Linoleic (C18:2) 0.6 0.71

2. Problem identification and looking for solutions El-Hadidi et al. 1991 described the conservation status of M. peregrina as “vulnerable” in the national Red List of Egypt (IUCN, 1992). After a quarter of a century, its status became worse. Numerous threats affect the existence of the tree M. peregrina in Sinai, Egypt; such as climate change; extreme drought for several years (Fig. 5) und unexpected flash floods (Dadamouny and Schnittler, 2016), over- and collection for different uses that led to a poor conservation status (Zaghloul, Dadamouny and Moustafa, 2012, Dadamouny et al. 2016). 19

Genetic diversity and conservation biology of Moringa peregrina Introduction

The overaged populations are subjected as well to infections. These threats probably caused M. peregrina to become extremely rare in Sinai. We assume a severe reduction in size for the remaining populations. Besides, an erosion of genetic diversity is likely to occur. The unmanaged utilization of this valuable tree is very short- sighted and led to high mortality, low recruitment, and poor survival of seedlings. Dadamouny (2009) found a non-natural age structure, indicating unhealthy shrinking populations with a sharp decline in the last three/four decades and a high rate of mortality among both very young and very old trees. If the threat situation remains unchanged, the current populations will permanently disappear. Therefore, conservation efforts should be directed to M. peregrina trees in Sinai during the establishment and growth periods to avoid its extinction.

Fig. 5. Impact of climate change (extreme drought and temperature) on the survival of M. peregrina at W. Feiran in South Sinai, Egypt.

Genetic diversity is of fundamental importance in the continuity of a species as it provides the necessary adaptation to the prevailing biotic and abiotic environmental conditions, and enables change in the genetic composition to cope with changes in the environment (Aikpokpodion, 2012). The combination of the proposed genetic analyses coupled with detailed studies of the population ecology of this species, will also greatly assist the design of cultivation practices for M. peregrina to preserve existing genetic diversity while enhancing the selection of desirable economic traits (e.g. high pod or oil yield). Conservation, sustainable use, rehabilitation, and bringing this tree species into cultivation will raise the poverty level of the local Bedouins people and will improve the health of their communities.

The investigations about genetic diversity of the remaining populations and their degree of heterozygosity of its populations were need. Therefore, the following study will help in answering many questions related to the genetic diversity within M. peregrina populations like; are cohorts of old trees genetically more diverse than cohorts of young ones? Is there a loss of genetic diversity? Is there any gene flow between the remaining M. peregrina populations or are these isolated? As a result for solving these problems, the following study will help to assess both degree and distribution of genetic diversity. Also, great efforts are needed to select genotypes with desirable features (high pod weight and/or oil content) for cultivation, and to develop in situ and ex situ conservation strategies for M. peregrina in Sinai. In the following study, we aim to study the genetic diversity and heterozygosity as well as gene flow between populations, to detect signs of possible genetic depletion and derive a conservation plan for the species. 20

Introduction Genetic diversity and conservation biology of Moringa peregrina

3. Goals of the thesis The following study mainly aims to: 1. Evaluating threats affecting the dynamics of the remaining M. peregrina populations in Sinai, Egypt. 2. Study the population performance of M. peregrina growing in Sinai, Egypt. 3. Evaluating its degree of endangerment and monitor the conservation status of the species based on the IUCN categories for threatened species. 4. Study the levels and distribution of genetic variation within M. peregrina populations. 5. Identify the relationship between Moringa tree performance, population size, and variation in fitness characters.

21

Genetic diversity and conservation biology of Moringa peregrina Introduction

References Abd El-Wahab, R.H. (1995). Reproduction Ecology of Wild Trees and in Southern Sinai, Egypt. M.Sc. Thesis, Botany Department, Faculty of Science, Canal University, Ismailia, Egypt. 110 pp. Aikpokpodion, P.O. (2012). Defining Genetic Diversity in the Chocolate Tree, Theobroma cacao L. Grown in West and Central Africa. In Genetic Diversity in Plants, edited by Mahmut Calişkan. InTech, Croatia, 498 pp. Al-Gohary, I.H., & Hajar, A.S. (1996). On the Ecology of Moringa peregrina (Forssk.) Fiori Anatomical Responses to Varying Soil Moisture Contents. Science, 8(1). Al-Kahtani, H.A. & Abou-Arab, A.A. (1993). Comparison of physical, chemical, and functional properties of Moringa peregrina (Al-Yassar or Al-Ban) and soybean proteins. Journal of Cereal Chemistry, 70(6): 619-626 pp. Al-Kahtani, H.A. & Abou-Arab, A.A. (1993). Comparison of physical, chemical, and functional properties of Moringa peregrina (Al-Yassar or Al-Ban) and soybean proteins. Journal of Cereal Chemistry, 70(6): 619-626 pp. Ashfaq, M., Basra, S.M.A. AND Ashfaq, U., (2012). Moringa: A Miracle Plant for Agro-forestry. J. of Agric and Soc. Sc Vol. 8, pp. 115–122. Boulos, L. (1999). 'Flora of Egypt.' Vol. I (Azollaceae – Oxalidaceae). Al hadara Publishing, , Egypt, 419 pp. Carlquist, S. (1988). Comparative wood anatomy - systematic, ecological and evoluti onary aspects of dicotyledon wood: Sprin ger Verlag, Berlin, New York. London: 216-238 pp. Cronquist, A. (1981). An integrated system of classifcation of flowering plants. New York: Columbia University Press. Dadamouny M.A., Unterseher M., König P., Schnittler M. (2016). Population performance of Moringa peregrina (Forssk.) Fiori (Moringaceae) at Sinai Peninsula, Egypt in the last decades: consequences for its conservation. J for Nature Conservation 34:65–74 http://dx.doi.org/10.1016/j.jnc.2016.08.005 Dadamouny, M. A., Zaghloul, M. S., & Moustafa, A. A. (2012). Impact of Improved Soil Properties on Establishment of Moringa peregrina Seedlings and a Trial to Decrease its Mortality Rate. Egyptian J. Botany, 52(1), 83–98. Dadamouny, M.A. (2009). Population Ecology of Moringa peregrina growing in Southern Sinai, Egypt. M.Sc. thesis, Environmental Sciences, Bot. Dept., Faculty of Science, Suez Canal University, Ismailia, Egypt. p 205. Dadamouny, M.A., & Schnittler, M. (2016). Trends of climate with rapid change in Sinai, Egypt. J. of water and climate change, 7(2), 393–414. IWA doi: 10.2166/wcc.2015.215. Dahlgren, G. (1989). The last Dahlgrenogram. System of classification of the dicotyledons. In: Tan K, ed. The Davis and Hedge Festschrift. Plant , phytogeography and related subjects. Edinburgh: Edinburgh University Press, 249–261. Dahlgren, R. (1980). A revised system of classification of the angiosperms. Botanical Journal of the Linnean Society 80: 91±124. El-Hadidi, M. N., Batanouny, K.H., & Fahmy, A.G. (1991). The Egyptian Plant Red Data Book, Volume I, Faculty of Science, , United Nations Environment Programme (UNEP) Geneve and Nairobi, p. 226. El-Lamey, T.M. (2015). Ecophysiological responses of Moringa peregrina (Forssk.) Fiori growing naturally under different conditions of Eastern Desert and Fieran Oasis, Egypt. Journal of Agriculture and Veterinary Science, Vol 8(4), 8-21. Freiberger, C.E., Vanderjagt, D.J., Pastuszyn, A., Glew, R.S.; Mounkaila, G., Millson, M., Glew, R.H. (1998). Nutrient content of the edible leaves of seven wild plants from Niger. Plant Foods for Hum. Journal of Nutrition, 53: 57-69 pp. Hartwell, J.L. (1971). Plants used against cancer: a survey. Lloydia's Journal, 34(1):103-1160 pp. Hindi, S.S. (2017). Some Promising Hardwoods for Cellulose Production: I. Chemical and Anatomical Features. Nanoscience and Nanotechnology, 4(3), 86-97. Kassas, M. & Zahran, M.A. (1962). Studies on the Red Sea costal land II. The district of El-Qibliya to . Journal of Ball-Social Geography, Egypt 38: 155-193 pp. Kassas, M., Zahran, M.A., (1971). Plant life on the coastal mountains of the Red Sea, Egypt. Journal of Indian Botanical Society 50A, 571-589 pp. Melchior, H. (1964). Adolf Engler Ed. Syllabus der Pflanzenfamilien, 2nd ed., II (Band). Olson, M. E. (2002). Intergeneric relationships within the Caricaceae-Moringaceae clade (Brassicales) and potential morphological synapomorphies of the clade and its families. International Journal of Plant Sciences, 163(1), 51-65. Olson, M.E. (1999). A wood monograph of the Moringaceae. Haseltonia 8. Olson, M.E. (2002a). Intergeneric relationships within the Caricaceae-Moringaceae clade (Brassicales) and potential morphological synapomorphies of the clade and its families. International Journal of Plant Sciences 163: 51-65. Olson, M.E. (2002b). Combining data from DNA sequences and morphology for a phylogeny of Moringaceae (Brassicales). Systematic Botany; 27: 55-73.

22

Introduction Genetic diversity and conservation biology of Moringa peregrina

Olson, M.E., Carlquist, S. (2001). Stem and root anatomical correlations with life form diversity, ecology, and systematics in Moringa (Moringaceae). Botanical Journal of the Linnean Society, 135: 315–348. Pax F. (1936). Moringaceae, Bretschneideraceae. In: Engler A, Prantl K, eds. Die natuX rlichen P¯anzenfamilien 17b. Leipzig: Engelmann, 693±698, 699±700. Pax F. 1891. Moringaceae. In: Engler A, Prantl K, eds. Die natuX rlichen Pfanzenfamilien III, 2. Leipzig: Engelmann, 242±244. Price, M.L. (1985). Moringa Tree. ECHO Technical Note. Educational Concerns for Hunger Organization, N. Ft. Meyers, FL. http://www.echotech.org/technical/technotes/ Quattrochi, U. (2000). CRC World Dictionary of Plant Names: Common Names Scientific Names, Eponyms, Synonyms, and Etymology (Vol 3), CRC Press, Washington, ISBN: 0849326737, 9780849326738, 887 pp. Quattrochi, U. (2000). CRC World Dictionary of Plant Names: Common Names Scientific Names, Eponyms, Synonyms, and Etymology (Vol 3), CRC Press, Washington, 887 pp Rodman JE, Price RA, Karol KG, Conti E, Sytsma KJ, Palmer JD. (1993). Nucleotide sequences of the rbcL gene indicate monophyly of mustard oil plants. Annals of the Missouri Botanical Garden 80: 686±699. Ronse Decraene, L.P., De Laet, J., & Smets, E.F. (1998). Floral development and anatomy of Moringa oleifera (Moringaceae): what is the evidence for a capparalean or sapindalean affinity?. Annals of Botany, 82(3), 273-284. Somali, M.A., Bajneid M.A., & Al-Fhaimani S.S. (1984). Chemical composition and characteristics of Moringa peregrina seeds and seeds oil, Journal of the American Oil Chemists' Society, 61(1): 85-86 pp. Täckholm, V. (1974). Student’s Flora of Egypt, Cairo University, Cairo, 1180 pp. Takhtajan, A. (1997). Diversity and classification of flowering plants. New York: Columbia University Press. Tsaknis, J. (1998). Characterisation of Moringa peregrina Arabia seed oil. Grasas y aceites, 49(2), 170-176. Zaghloul MS, Moustafa AA, Dadamouny MA (2012). Age structure and population dynamics of Moringa peregrina, an economically valuable medicinal plant. Egyptian J Botany 52(2):499–519. Zahran, M.A. &. Willis, A.J. (2009). 'The vegetation of Egypt'. 2nd Ed., London: Chapman and Hall, 437 pp.

23

Genetic diversity and conservation biology of Moringa peregrina 3 populations in Sinai, Egypt Publications

Original Article (published): 3.1. Trends of climate with a rapid change in Sinai, Egypt. 25 Original Article (published): 3.2. Population performance of Moringa peregrina (Forssk.) Fiori (Moringaceae) at Sinai 52 Peninsula, Egypt. Original Article (Manuscript): 3.3. Demographic and size structure status of the threatened populations of Moringa 78 peregrina trees at Sinai Peninsula, Egypt. Original Article (Manuscript): 3.4. Relictic populations of the endangered tree Moringa peregrina (Moringaceae) in Sinai, 96 Egypt are still genetically diverse.

24

Trends of climate with a rapid change in Sinai, Egypt Publications 3.1 Trends of climate with a rapid change in Sinai, Egypt

Mohamed A. Dadamouny and Martin Schnittler Ernst-Moritz-Arndt University of Greifswald, Institute of Botany and Landscape Ecology, Soldmannstr. 15, D-17487 Greifswald, Fax +49 03834 420 4114 *Email: [email protected]

Abstract This study presents evidence for rapid climate change in the Sinai Peninsula, Egypt. Analyses of data for temperature and rainfall from 1970 to 2014 show a clear tendency towards decreasing rainfall and increasing average temperatures. This trend caused severe droughts for many years that were suddenly interrupted by high and unpredictable rainfall that fluctuated heavily in space and time. If this tendency continues, the population dynamics of many plant and animal species will be negatively affected, with many of them being important for local inhabitants. Detrimental effects can be expected in the coastal and tourist cities like Sharm El-Sheikh, Taba, El-Tor, St. Catherine, Ras Sedr and El-. Conservation efforts should be directed to conserve the biological and natural resources and to keep pace with this environmental change. Key words: aridity, climate, conservation, resources, Sinai, trends

1. Introduction The world’s population is anticipated to increase from around 7.2 to 9.2 billion (109) within the period from 2014–2050. The monthly growth is estimated to be around 6 million (Lal and Stewart 2012). Most of it occurs in developing countries, where natural and biological resources are currently under considerable human pressure (El-Ramady et al. 2013). For this reason, changing climate affects human lives, particularly severely in these regions of the world. However, climate change is a complex phenomenon not easy to circumscribe. Commonly measured by meteorologists are daily weather data, air temperature, precipitation, atmospheric pressure, humidity, wind speed, sunshine, and cloud cover (Food and Agriculture Organization [FAO] 2008). Obviously, these parameters are not stable and can change within short periods of time. Scientists have not yet agreed upon an internationally accepted definition of climate change. According to FAO (2008), it can refer to (1) long-term changes in average weather variables (World Meteorological Organization); (2) All changes in the climate system, involving the drivers of change and their impacts (Global Climate Observing System); and (3) Just man- induced changes in the climate system (United Nations Framework Convention on Climate Change) (El- Ramady et al. 2013). This paper is describing climate trends at the Sinai Peninsula, Egypt. We will adopt the first definition since it is extremely difficult (and not the focus of this paper) to disentangle the drivers of climate change, especially if these drivers are natural processes or stem from human impacts. Only some causal

explanations are well known. Irrespective of its prediction around 120 years ago by scientists studying CO2, man-made climate change was formally approached first in 2007 by the Nobel Prize Committee (Lichtfouse 2009). Already Svante Arrhenius (1859–1927) estimated that the greenhouse impact, by doubling the

25

Publications Trends of climate with a rapid change in Sinai, Egypt

concentration of the atmospheric CO2, would cause a temperature rise of about +5 to +6°C (Arrhenius 1896). Although his crude estimate is higher than most of today’s scenarios, he is considered the first scientist to have predicted anthropogenic global warming (Arrhenius 1908, see Weart 2008). We know today that his pathway is one of many possible drivers of the climate. It included natural drivers (e.g. Peatlands, where a significant part of the carbon held by the vegetation returns to the atmosphere due to plant and microbial respiration,

Gorham 1991; Lappalainen 1996). Also, it included anthropogenic (Kyoto Protocol) processes. Besides CO2,

methane (CH4), nitrous oxide (N2O), hydrofluorocarbons (HFC), perfluorocarbons (PFC), and sulfur

hexafluoride (SF6) are greenhouse gases emitted by human activities (Chanton et al. 1995). Consequently, the International Panel on Climate Change (IPCC 2014, 5th Assessment Report) presented Representative Concentration Pathways (RCPs) referring to more than one possible scenario for future climate change. RCPs were calculated up to the year 2100; with Extended Concentration Pathways (ECPs) describing extensions of the RCPs from 2100 to 2500. The increase of global mean surface temperature (2081–2100) relative to (1986– 2005) is expected to be 0.3 to 1.7°C under RCP2.6, 1.1 to 2.6°C under RCP4.5, 1.4°C to 3.1°C under RCP6.0 and 2.6°C to 4.8°C under RCP8.5.

Egypt’s contribution to global emissions of greenhouse gases is rather small (3.4 % coal/peat in 2011) (Egyptian Environmental Affairs Agency [EEAA] 1999; International Energy Agency [IEA] 2013). But, the potential impacts (social and economic) of climate change could be harsh for the country’s future (EEAA 1999; Comsan 2011). These effects are due to the steadily growing population of Egypt, its large proportion of deserts and limited fertile land, and the concentration of economic activities near the coastline. Egypt’s climate is semi- arid and characterized by hot dry summers, moderate winters, and very little rainfall (Dadamouny 2009). Therefore, most of the land can be classified as semi-desert or desert. The impacts of climate change, in particular warming, are reflected not only by changes in the water cycle (Howard et al. 2010). Climate change has also led to increasing desertification and higher mortality in many plant species that cannot tolerate high temperature and severe droughts (Figure S1a-c). Other risks are less predictable, including flash floods which may lead to unpredictable water flows and recharge (Figure S1d), more frequent droughts, and changes in the capacity and nature of groundwater stores (Stern 2006; Xu et al. 2007; Bates et al. 2008). Also, a rise in sea level will increase the risks of permanent or seasonal saline intrusion into the groundwater in South Sinai and the River in its northern delta (Howard et al. 2010). This is affecting the quality and potential usability of water sources for domestic, agricultural and industrial uses. Consequences of climate change will impede all development sectors in Egypt. The most significant repercussions include impacts on the water resources of the Nile River and groundwater, agricultural output, coastal resources, and tourism. As the environmental conditions play a crucial role in defining the function and distribution of species, changes in these conditions will actively change plant diversity patterns in the future (Duraiappah 2006; Field et al. 2009). In the upcoming years, it is expected that climate change will be one of the main drivers of biodiversity patterns (Sala et al. 2000; Duraiappah 2006; Pressey et al. 2007; Ingole and Kakde 2013).

In this paper, we try to derive climate trends for the Sinai Peninsula. This region is located between the and the Red Sea, including around 2000 plant species and many hundreds of rare and 26

Trends of climate with a rapid change in Sinai, Egypt Publications

endemic animals. Based on data available from weather stations recorded from 1970 to 2014, we calculate trends and briefly discuss their impacts on the status of natural and biological resources. Finally, we present some proposed solutions that may help to mitigate effects of climate change in the near future.

2. Study Area Sinai lies in the arid to the extremely-arid belt of and belongs to the Saharan-Mediterranean climate region. Most of Sinai (excluding the mountains) is classified as extremely arid (P/EP< 0.03), where P is the annual precipitation and EP is the potential evapotranspiration, calculated according to Penman's formula (Ayyad and Ghabbour 1986). The mountains in the St. Catherine area, with their highest summits above 2,600 m a.s.l., receive higher amounts of precipitation. It is estimated to exceed 50 mm a year as rain and snow (Dadamouny 2009). We selected six regions for an analysis of climatic parameters (Figure 1). The first region is El-Arish, which is the capital and largest city of North Sinai. It lies on the Mediterranean coast of the Sinai Peninsula (31°08'N, 33°83'E) about 50 kilometers from the Rafah border with the Gaza Strip and 344 kilometers north-east of Cairo. Most prominent is a broad wadi known as W. El-Arish, which receives flash flood waters from north and central Sinai. Although the prevailing Mediterranean winds alleviate its high temperatures, as for the northern coast of Egypt, the region is classified as a hot desert. Its basin area which drains into the Mediterranean Sea, covers about 19,500 km2 (El-Bihary and Lachmar 1994; Klein 2000, Fig. S2a). The coastline is rather straight, receiving sediment loads varying from 500,000 to 800,000 m3/year, and the coastal sediment bar usually dams the Nile River outlet (Nir 1982, 1989).

EMA in Cairo

Fig. 1. Map of the Sinai Peninsula, Egypt, showing the locations of the weather stations at 1. El-Arish, 2. Ras Sedr, 3. St. Catherine, 4. Sharm El-Sheikh, 5. El-Tor, and 6. Taba.

27

Publications Trends of climate with a rapid change in Sinai, Egypt

The second region is Ras Sedr, known as well by the name “Sudr”. It is located on the coast (29°35'N, 32°42'E) and belongs to the South Sinai Province which consists of three areas: W. Sedr, Abu Sedr, and Soerp. Ras Sedr as a touristic city has a 95 km beach coastline. It lies 200 km from Cairo and approximately 60 km from the Ahmed Hamdi Tunnel crossing the Suez Canal. In the Köppen-Geiger climate classification system, the area is considered as a hot desert (Peel et al. 2007). It is also highly vulnerable to flash floods and drought that may be attributed to climate changes (Ahmed et al. 2014).

The third region is St. Catherine at the central part of South Sinai (28°34'N, 33°57'E). This mountainous region of Precambrian igneous and metamorphic rocks includes Egypt’s highest peaks (Mt. St. Catherine, 2,641 m a.s.l., Mt. Moussa 2,285 m, Mt. Serbal 2,070 m., Mt. Umm Shomer 2,586 m, and Mt. Tarbush 2,029 m). The Sinai massif contains some of the world’s oldest rocks and around 80% of these rocks are 600 million years old (Moustafa 1990; Said 1990; Dadamouny 2009). According to the Köppen-Geiger climate system, its climate is also a hot desert (Peel et al. 2007). However, St. Catherine city has the coldest nights in Egypt due to its high elevation (1,500 m a.s.l.); minimum temperatures can reach -15°C. Among its high mountains, there are basins and many wadis (W. Al-Alarbaeen, Shrayj, Ferrah, Itlah, Raha, Tellaah, Naqb Hawa, Farsh El-Roumana, Farsh Mesaila, Safsafa, Tynia, Ahmar, Farsh El-Luza, Gebal, and W. El-Deir). Due to the rugged relief, weather conditions change locally and within a short time, including extreme weather events with snows on the summits. The region comprises two main wadis; W. Zaghra situated north-east of St. Catherine (Fig. S2b) and W. Watir (Fig. S2c) near . The latter wadi is among the wadis in the Sinai hit most often by flash floods (Fig. S2c), with a catchment of 3,580 km2, which is considered to be a hyper-arid catchment (Lin 1999). A third wadi, W. Feiran, is the longest and broadest wadi in South Sinai. Its basin surface is 1,675 km2 and covers a densely branched and widely distributed drainage system.

The fourth region is Sharm El-Sheikh at the southern tip of the Sinai Peninsula. It lies on the coastal strip along the Red Sea overlooking the at the mouth of the Gulf of Aqaba (27°55'N, 34°19'E). Its climate is a subtropical arid one. The Gulf of Aqaba coast is warmer than the Gulf of Suez (Abd El-Wahab 2003). The coastlines of the Red Sea and outlying reefs are increased at irregular intervals by creeks known as Sharm, which are typically drowned stream valleys. Near to Sharm El-Sheikh are two more wadis (W. Mandar and W. Lithi) and the National Park Ras Mohamed (480 km2), situated around 30 kilometers south of Sharm El-Sheikh.

The fifth region is El-Tor at the Gulf of Suez, South Sinai (28°14'N, 33°37'E, Dadamouny 2009), known as well under the name Tur Sinai. A prominent wadi is W. Meir, situated north-east of El-Tor. The sixth region is Taba. It is a small Egyptian town covering 1 km2 near to the northern tip of the Gulf of Aqaba (29°29'N, 34°54'E), its protectorate south-west of Taba covers 3,590 km². The area within 40 km of its meteorological station is a hot desert (Peel et. al., 2007) covered by shrub-lands (68%), oceans and seas (18%), and lakes and rivers (13%, Shokry and Ammar 2009).

28

Trends of climate with a rapid change in Sinai, Egypt Publications

3. Material and Methods 3.1. Data collection Available meteorological data for the selected regions in Sinai were collected from six meteorological stations in June 2014. These stations are El-Arish, Ras Sedr, St. Catherine, Sharm El-Sheikh, El-Tor, and Taba, all governed by the Egyptian Meteorological Authority (EMA, established 1898 in Cairo). Later, the data were updated up to December 2014. We collected the daily raw data for the parameters: minimum, maximum and mean temperatures, total precipitation, relative humidity, air pressure at sea level, and wind speed. A quality control has been set up for the studied stations by EMA since 1970. Data are double controlled, first by the observer and second by a supervisor who makes occasional visits at each station, checking records at different times of the day and repeating measurements. Results are submitted to EMA. 3.2. Data treatment From 1970 to 2014, all years with daily records for less than eight months were considered as missing. For these years, we took the monthly and annual means from the unpublished data books prepared by the Egyptian Meteorological Authority (EMA). They prepared two books; the first includes data from 1945 to 1974 but has only one of the six studied stations (El-Tor). The second book contains the monthly and annual means from 1975 to 2005 for all locations. For St. Catherine and Sharm El-Sheikh, there were two stations in each location, but the data in one of them were missing for many years, therefore we considered only one station in our analysis. We completed the missing years not available in the stations’ reports from EMA in Cairo. For the temperature, the missing data totaled 6.8% (37 out of 540 months). For the rainfall, the data of some years appear to be missing, but in certain cases this is because there were no rains. The missing rainfall data ranged from 4.4% to 13.3% (two, five, and six years were missed for El-Arish, Sharm El-Sheikh, and El-Tor and St. Catherine) respectively. The availability of data for Taba station was 99.2%. Rainfall records for Ras Sedr station before 1978 were also missing. To verify the equality of variance (P> 0.05) in the data (homogeneity), a non-parametric Leven’s test was used according to Nordstokke and Zumbo (2010) and Nordstokke et al. (2011) using IBM SPSS Statistics version 22 (SPSS, 2013). 3.3. Data analysis First, from the daily records we calculated the monthly and annual means, minimum, maximum, standard deviation (SD), and standard error (SE) values for all collected parameters through the descriptive statistics. Based on the analyses of climate records from 1970 to 2014, a linear regression equation between X (years) and Y (climate parameters) was used to forecast the trend of these parameters within future years. We used the software Minitab 17 and SPSS 2013. For temperatures, alternatively, the pattern was visualized by the change in monthly means for the time intervals 1970–1984, 1985–1999, and 2000–2014. The total number of rainy days per year, a percentage of the highest three days to the total annual rainfall, highest daily records, and total annual rainfall were used to draw the trend of precipitation in the study area. Regression analysis was carried out for yearly minimum, mean, and maximum values to reveal the extremes and to forecast the tendency of each. For the same period (1970–2014), regression of the total rainfall was carried out for the studied six regions except Ras Sedr (from 1978 to 2014), due to the missing data for the rainfall before 1978. Subsequently,

29

Publications Trends of climate with a rapid change in Sinai, Egypt

regression for annual means of relative humidity, air pressure at sea level, and mean and maximum wind speed was carried out using the available climate data for the six regions from 1970 to 2014.

4. Results 4.1. Changes of climate parameters over time Table 1 lists the figures for linear trends calculated from annual records of minimum, maximum and mean temperatures, annual precipitation, relative humidity, air pressure and maximum and mean wind speed. Table S1 gives the respective data for the years 1995 to 2014. 4.1.1. Temperature From 1970 to 2014, the average of annual mean temperature of the studied six regions in Sinai increased by +2.3°C (20.3 to 22.6°C). Figure 2 displays the increase of monthly mean temperature during the timespans 1970–1984, 1985–1999, and 2000–2014 in each region. From the first time interval (1970–1984) to the last (2000–2014), the highest increase of the annual mean temperature (+2.8°C) was at Taba on the Gulf of Aqaba. While, the lowest increase in the annual mean (+1.2°C) was in El-Arish on the Mediterranean Sea. The extreme values of maximum and minimum temperatures are shown in Fig. 3. During the last 20 years (1995–2014), the maximum recorded temperatures were 46°C for Sharm El-Sheikh (22 and 23 June 2010 then 2 and 3 June 2013). Then 45°C was recorded at El-Arish (29 May 2003, 25 June 2007, and 4 Mar. 2013), Ras Sedr (30 July 2002), El-Tor (2 July 1995 and 20 August 2010), and Taba (2 July 1996, Table S1). Absolute minimum temperatures were -6°C, recorded at El-Arish on 8 Jan. 1994, and at St. Catherine on 1 Feb. 2008 (Fig. 3). Moreover, Sharm El-Sheikh is the warmest region in our study area, its minimum temperature was 4°C on 5 Feb. 2003 and at El- Tor it was 1°C on 1 Jan. 2003.

Maximum temperatures tend to rise in all studied regions (Fig. 2). The regression coefficients for the maximum and mean temperatures in all stations are positive. If this trend continues in a linear fashion, maximum temperatures may reach 52°C at El-Arish, 53°C at Ras Sedr, 49°C at St. Catherine, 54°C at Sharm El-Sheikh, 55°C at El-Tor, and 46°C at Taba by 2050. At the same time, the mean temperature is increasing and may exceed 30°C at Sharm El-Sheikh. The changes are much less pronounced for minimum temperatures; which tend to rise in all studied regions except for St. Catherine, where the regression coefficient is low (0.24, Table 1). According to the linear trends, minimum temperatures may reach 15°C at Sharm El-Sheikh, 10°C at El-Tor, 8°C at Ras Sedr and Taba, and 2°C El-Arish by 2050. In St. Catherine, it may be lower than -3°C in the upcoming years. A stronger fluctuation in temperature is one indication of climate change. Extreme temperatures seem to occur more often in recent years, especially for maximum temperatures (Fig. 3). 4.1.2. Annual total precipitation Data for total precipitation (TP) from 1970 to 2014 are shown in Fig. 3. Single, comparatively wet years with rainfall peaks were 1975 (271.9 mm), 2002 (205.5 mm), and 2008 (142.7 mm) at El-Arish, 1978 (64.4 mm), 1987 (41.1 mm), and 2006 (94.5 mm) at Ras Sedr, 1975 (97.5 mm), 76 (126.5 mm) 91 (68 mm), and 2010 (44.1 mm) at St. Catherine, 1975 (48.3 mm), 90 (19 mm), and 2010 (34.4 mm) at Sharm El-Sheikh, 1975 (47.5 mm), 2002 (82.6 mm), and 2011 (18.6) at El-Tor, and 1978 (150 mm), 88 (51.6), and 2010 (61.5 mm) at 30

Publications Trends of climate with a rapid change in Sinai, Egypt

Table 1. Intercept (a) and slope (b) of a linear regression as well as the regression coefficient (R2) for eight climate parameters (maximum, mean, and minimum temperature; annual precipitation, mean relative humidity, mean pressure at sea level, and maximum and mean wind speed) measured at six stations of the Sinai peninsula (1970–2014).

Max. temperature (oC) Mean temperature (oC) Min. temperature (oC) Total annual rainfall (mm)* a b R2 a b R2 a b R2 a b R2 El-Arish - 445.1 0.243 0.71 - 54.5 0.038 0.32 - 32.4 0.017 0.02 6027.9 - 2.96 0.38 - 462.4 0.251 0.77 - 129.4 0.076 0.62 - 0.073 0.27 657.4 - 0.319 0.04 Ras Sedr* 142.8 St. Catherine - 602.9 0.318 0.66 - 118.1 0.068 0.42 125.3 - 0.063 0.24 2280.7 - 1.128 0.31 - 477.9 0.259 085 - 145.6 0086 0.75 - 0.161 0.53 736.7 - 0.364 0.14 Sharm 314.2 - 539.4 0.289 0.79 - 152.7 0.088 0.43 - 0.102 0.41 431.2 - 0.210 0.03 El-Tor 199.5 - 151.9 0.096 0.27 - 171.1 0.096 0.36 - 0.107 0.22 1852 - 0916 0.14 Taba 210.9

Mean RH (%) Mean sea level pressure (hPa) Max. wind speed (km/hr) Mean wind speed (km/hr) a b R2 a b R2 a b R2 a b R2 El-Arish - 17.2 0.042 0.06 1,792.0 - 0.393 0.02 370.5 - 0.157 0.03 73.0 - 0.033 0.05 Ras Sedr* 103.4 - 0.026 0.02 1,374.0 - 0.180 0.65 207.9 - 0.082 0.02 44.2 - 0.016 0.03 - 1,112.2 0.5721 0.30 3,155.6 - 1.068 0.103 - 0.505 0.01 154.4 - 0.068 0.08 St. Catherine 916.5 Sharm 640.8 - 0.300 0.37 1,514.1 - 0.251 0.86 1,841.8 - 0.884 0.12 632.5 - 0.306 0.65 El-Tor 550.7 - 0.247 0.20 1,134.9 - 0.062 0.12 1,490.8 - 0.702 0.06 98.4 - 0.037 0.03 Taba 46.5 - 0.268 0.20 1,014.1 - 0.075 0.18 2450 - 1.193 0.48 412.4 - 0.198 0.26

* Regression for total rainfall at Ras Sedr is based on data from 1978 to 2014.

31

Publications Trends of climate with a rapid change in Sinai, Egypt

Year

C) o

Meantemperature Monthly (

C) o

( temperature Annual Mean

Fig. 2. Trend of mean temperatures (°C) in the studied six regions based on monthly means between 1970 and 2014 (lower left) in three time intervals (each 15 yrs). Long term regression (upper right) is for the annual means Period 1970 - 2014.

Taba. However, like the average amounts of precipitation, these peaks tend as well to decrease. The average of total annual rainfall±SE (1970–2014) was 119.8±9.7 (El-Arish), 35.3±4 (St. Catherine), 10.8±2 (Sharm El- Sheikh), 13±2.6 (El-Tor), 26.6±4.7 (Taba), and 19.7±3 (Ras Sedr, 1978–2014). Within the time span 1970 to 2014, the most striking trend is the severe decline of precipitation (both rain and snow, Fig. 4). Based on the analysis of total rainfall during this period, the trend of rainfall is different from that recorded for temperature. Especially, severe declines were recorded for South Sinai, except for one or two wet years (Fig. 4 and S3). The

32 Trends of climate with a rapid change in Sinai, Egypt Publications

Figure 3. Extreme values of absolute maximum and minimum temperature and total annual rainfall from 1970 to 2014.

slopes of the regression equations and their coefficients are shown in Table 1. The latter values are negative, indicating the decline of total rainfall year after year. On 19 and 20 February 1975, an extreme rainfall event occurred in the El-Arish drainage basin, causing an extreme flood flow throughout the Sinai Peninsula except for its northern part. Table S1 shows the total annual rainfall during the last 20 years (1995–2014). Years with unusually high precipitation stick out, like 205.5 mm measured in 2002 at El-Arish, which is located on the 33 Publications Trends of climate with a rapid change in Sinai, Egypt

Mediterranean Sea. This is followed by Ras Sedr, where 94.5 mm was recorded in 2006 or 82.6 mm recorded at El-Tor during 2002. High annual rainfalls were recorded as well for St. Catherine (44.1 mm) and Sharm El- Sheikh (34.4 mm, both in 2010, Table S1). Based on analyses of daily precipitation, maximum daily figures during the period from 1995 to 2014 were 91.9 mm (18 Feb. 2002, El-Arish), followed by 86.1 mm (1 Apr. 2006, Ras Sedr) and 70.1 mm (3 Sep. 2002, El-Tor).

Moreover, our available few years before 1970 showed that the precipitation was distributed in the months of winter and spring (October to March or April). But, during the years from 1970 to 2014, we reported a change in rainfall pattern as the maximum daily amounts of rainfall are received in one to three days of the year. Figure 4 shows the total number of rainy days. Previously, in El-Arish, the total annual rainfall in 1963 was 163.58 mm recorded in 20 days, and the highest daily rainfall was 36.07 mm in December. At the same station, the total rainfall/ number of rainy days/ maximum daily record were 117.87 mm/ 27/ 41.91 mm (21 April) in 1964, 171.70 mm/ 23/ 35.05 (12 Jan.) in 1965, and 77.98/ 22/ 20.7 (23 Mar.) in 1966.

4.1.3. Relative humidity Trend for annuals means of relative humidity (RH) declined for all regions except in two stations. These stations were El-Arish, which shows a slight increase and St. Catherine, which is the coldest region in Egypt due its high elevation (Table 1). Highest annual means of relative humidity were recorded at Al-Arish in 2006 (76%, Table S1), followed by 65.3% at Taba in the same year. The annual means range between 49% and 58% at Ras Sedr, 21% and 43% at St. Catherine, 35% and 49% at Sharm El-Sheikh, and 43% and 59 % at El-Tor as shown in Table S1.

4.1.4. Air pressure at sea level With increasing the temperature, the mean air pressure at sea level (PSL) seems to decline (Table 1). During the investigation period (1970–2014), the lowest/highest annual means were around 843/1017 hectopascals (hPa) at El-Arish. They were 1011/1016 at Ras Sedr, 934/1022 at St. Catherine, 1009/1015 at Sharm El-Sheikh, 1009/1012 at El-Tor, and 1015/1022 at Taba (Table S1).

4.1.5. Wind speed Increasing annual temperatures seem to be followed by a decline of maximum wind speed. Maximum figures (226 km/h) were recorded on 29 Dec. 2010 at St. Catherine, followed by 115 km/h at El-Tor on 6 June 1996, then 108 km/h at Sharm El-Sheikh on 21 July 2000. The extremes of the wind speed are usually accompanied by rains with sudden gusts in these days. The maximum wind speed was 85 km/h at El-Arish (23 Feb. 2000). This is followed by 70 km/h at Taba (21 Apr. 2004), and 57 km/h at Ras Sedr (7 Mar. 1996). Much lower maxima were recorded for El-Tor (25.7 km/h), Sharm El-Sheikh (22.1 km/h), and Taba (21.6 km/h, Table S1).

34 Trends of climate with a rapid change in Sinai, Egypt Publications

Fig. 4. Trend of annual precipitation (mm) and maximum daily rainfall (mm) from 1970 to 2014 for the stations: El-Arish, St. Catherine, Sharm El-Sheikh, El-Tor, Taba, and from 1978 to 2014 for Ras Sedr station, change in the rainfall pattern through decrease in the number and amount of rainy days interrupted by the highest amount of rainfall in one to three days.

4.2. Impact of climate change on biological and natural resources Although we have data only for single species, the trends in temperature and rainfall are likely to have an adverse impact on vitality and population size of many rare plant and animal species in Sinai (Fig. S1). Often,

35 Publications Trends of climate with a rapid change in Sinai, Egypt

this is overlaid by anthropogenic impact, most prominently, overgrazing, the collection of fuelwood and fodder, and fire. Based on global positioning system (GPS) records, there are numerous coastal regions that may be flooded if water levels of the Mediterranean and the Red Sea rise in future. One of these regions is Nuweiba, located on the Red Sea coast (Fig. 5) at the mouth of W. Watir (Fig. S2d-g). Based on a satellite image, a possible increase of the Red Sea level (according to different scenarios anticipated as around 30 cm, 50 cm, and 1 m) may lead to the disappearance of Nuweiba Port and its surrounding vast floodplain (around 40 km2). Moreover, saline water may mix with groundwater and reach the water drainage of W. El-Arish (Fig. S2). Similar scenarios pose a threat to many coastal cities (Baltim, , Ras El-Bar, , El-Manzala, Borg El-Arab, and ). According to a worst-case scenario, the water may move through the northern delta upwards to reach Damanhour city, El-Mansoura, Shirbin, Kafr El-Sheikh, El-Mahalla El-Kubra, and .

Fig. 5. Satellite image for Nuweiba showing a coastal region that may be affected by an increase of sea level of the Red Sea and the flash flood of W. Watir.

5. Discussion Currently, the human contribution to climate change is a topic of intense debate among scientists. Most researchers believe that climate changes as well steadily due to natural processes, and untangling the complex natural and anthropogenic drivers of climate change is one of the most challenging tasks of future research. One

of the main questions is if the anthropogenic causes, like the CO2-mitigated greenhouse effect, will become the core drivers of future changes. Many other drivers of climate change should be considered, including the increase of the anthropogenic heat fluxes from cities (González-Aparicio et al. 2014) and energy balance on the earth. According to the IPCC (1990), an early study declares that the year of 1988 was the warmest year within 36 Trends of climate with a rapid change in Sinai, Egypt Publications

the last 100 years. Similarly, 1989 was the fifth warmest year in the 134 years with records of global temperature, and six of the ten warmest years on record occurred in the 1980s (Kerr 1990). Climatologists have different opinions on whether the observed increase in average annual global surface temperature (about +0.6°C) over the last century is already a clear indication of greenhouse warming or not (IPCC 1990). Otherwise, is it still within the natural variability of climate or not? There seems little dispute among climatologists that mankind has already significantly influenced the earth’s climate (Golitsyn 1989). This influence is through such factors as the urban heat-island impact, intensification of desertification through land degradation, deforestation and stratospheric ozone depletion (IPCC 1990). Some studies assume that global warming will become more severe at higher latitudes, affecting winters to a greater degree than summers (Golitsyn 1989; Schneider 1989). A warmer atmosphere contains more water vapor and upturns the intensity of the entire hydrological cycle, but precipitation trends are likely to change homogeneously in time and space (Golitsyn 1989). Widely discussed, but statistically even harder to ascertain, is the question of whether global warming will increase the likelihood of floods and droughts, and whether it causes stronger storms or typhoons, and severe heat waves (Golitsyn 1989; Hansen et al. 1989). 5.1. Trends in the climate of the Sinai region Climate change in our study area does not only affect water resources but may increase the aridity of the area as well by increased evaporation. It will increase desertification; arid and semi-arid environments already occupy around 37% of the earth’s surface (United Nations Environment Programme [UNEP] 1992). Around 64% of the arid regions and 97% of hyper-arid deserts are concentrated within Africa and Asia (Dadamouny 2009). The Sinai region, especially its southern part, is now extremely hyper-arid with a long hot and rainless summer and mild winters. Its climate is affected by the orographic and the tropical impacts of the mountains and the two gulfs (Suez and Aqaba) (see Migahid et al. 1959; Zohary 1973; Batanouny 1981; Issar and Gilad 1982; Danin 1983, 1986). Orographic precipitation predominates in southern Sinai, with clouds caught mainly by summits and cliffs. Gorges in the mountains help transport the water to upstream tributaries of the wadi system (Kassas and Girgis 1970). For the time of this analysis (1970–2014), we reported a trend of decreasing annual precipitation overlaid by extreme events causing unpredictable flash floods. Klein (2000) stated that extreme flood events may occur once in a century or even less frequently. We recorded the first severe flood in North Sinai at El-Arish in 1975 and this was repeated in 1987, 1998, 2002, and in 2008. In South Sinai, although the amount of rainfall is lower than in the north, few extreme days were recorded for the years 1975, 76, 79 and 84 in St. Catherine, or 2002 in El-Tor. During most rainy days, we observed a high wind speed or even storms. A typical front progresses from South Sinai to the northeast. In accordance, our results reveal extreme wind speeds in the southern part of Sinai; particularly around St. Catherine. A single rainstorm contributed the following amounts of rain on El-Arish drainage basin in 1975: 73 mm at St. Catherine Monastery (south of the basin), and 48 mm at Nakhl (central part of the basin, Gillad 1975). It is estimated that the average annual rain over the basin was 40–50 mm in this year. Although, the northern part of El-Arish basin did not receive significant precipitation (8 mm), the total volume of the flood flow in the basin was estimated at 8 x 108 m3 in 1975 (Smith et al. 1997; Klein 2000). North Sinai, especially the delta of Wadi El-Arish depends on groundwater as a core resource for drinking water, industrial, and agricultural use. Here, climate change may

37 Publications Trends of climate with a rapid change in Sinai, Egypt

have a direct impact on groundwater availability (Abdelaziz and Bakr 2012). Extreme rainfalls have been recorded as well in earlier years: Dames and Moore (1982) stated one rainy day delivering 76.2 mm in November 1937. But, the average rainfall during the last 30 years did not exceed 44.1 mm at St. Catherine. As such, extreme events contribute a significant amount of the annual precipitation. During the time interval of the current study (1970–2014), the extreme events became more frequent. In agreement with Jay and Naik (2011), the rainfall trend has decreased, and this behavior of rainfall pattern may be another indication of climate change. It seems to be a big part of the problem: a few days with extreme rains or flash floods and most of the water is lost instead of being absorbed by the soils. Abdel-Wahab (2003) and Dadamouny (2009) stated that rainfall of St. Catherine decreased from 60.4 mm in the 1930s to 42.6 mm in 1990s. But, at least the rain amount was distributed in the months of winter and spring. Thus, the trend of rainfall reflects a change in rainfall pattern, where the majority of the rainfall on Sinai, mostly in the south, is recorded in one to three days only.

In Sinai, increasing temperature coupled with drought are likely to affect plants, wild animals, and populations, similar to expected impacts of rising global temperatures on human health, lifestyles, food production, economic activity, residential and migration patterns (IPCC 1990). Combined with the associated general increase in temperature, changes in atmospheric circulation patterns and the water cycle may affect water availability, agricultural activity, flood protection practices, infrastructure planning, and natural habitats. Since, Sinai Peninsula is neither heavily industrialized nor populated, human influences are most likely to come from the outside, which invokes the full complexity of climate modeling to estimate human effects. Disturbance of the Sinai Peninsular groundwaters and drainages cycles or coastal erosion along the Red and Mediterranean Seas is feared. Many studies (such as Jay and Naik 2002, 2011) showed a rise in the range of tides throughout most of the eastern Pacific as a result of global climate change and human activities. Human interference through flood control of rivers, water withdrawal, hydropower generation, navigational development, and changes in land use led to changes in hydrologic trends (Naik and Jay 2011). Another fear is a higher frequency of extreme

weather events like droughts, floods, and tropical cyclones. Doubling the effective CO2 may increase the intensity of tropical cyclones or hurricanes by as much as 40% (Emanuel 1987). Further studies may provide a more reliable estimate of impacts of global warming on tropical cyclone activity (Henderson-Sellers and Blong 1989). Also, a significant rise in global mean sea levels can be expected from the thermal expansion of the upper layers of the ocean and meltdown of glaciers. IPCC scenarios forecast a rise of 9-29 cm over the next 40 years, or 28-98 cm by 2090 (IPCC 1990). An increase of only 25 cm or more in relative sea level would displace many residents of the delta regions of the Nile from their homes and livelihoods (Tickell 1989). This projected sea- level rise will also cause widespread coastal erosion, especially on shallow coasts (Titus 1988; Murday 1989). For a region like Sinai, where most human activities center around the coast and in a few big wadis, this might become a catastrophe.

For the average temperature trend, Sinai increased +2.3°C based on our analysis of data (1970–2014) (20.3 to 22.6°C). The same pattern is obtained by applying the General Circulation Model (GCM) at the Environment and Climate Changes Research Institute in Egypt on the Nile Delta (Equatorial plateau) but with different

38 Trends of climate with a rapid change in Sinai, Egypt Publications

scenarios. Change in temperature at the Nile Delta (2000–2100) is calculated with various models to be +3.2°C (CSIRO), +4.3°C (GFDL), +3.8°C (GISS), and +4°C (NCAR) (Abdel-Aziz 2011). As St. Catherine has a more changeable and unique climate than other regions of Sinai, a lot of studies were concerned with its climate. According to Dadamouny et al. (2012), during the period 1979–1992, the lowest monthly mean minimum temperature at St. Catherine ranged from 1.4 in Jan. to 17.5°C in July. St. Catherine and Nekhel are considered to be the coldest towns in Egypt, together with some settlements in mountainous Sinai. Snowfalls in St. Catherine take place regularly in the winter months (Dec.–Feb.) and occasionally in autumn and spring. From 1934 to1937, the lowest monthly mean temperature was recorded -15°C (, Dadamouny 2009). While the highest mean maximum temperature for the same period was 25.8°C (Aug.), it then increased during 1979–1992 to 30.8 and 31.8°C in June and July, respectively (Abd El-Wahab 2003), and to 34.3°C during 1970–2014. Moreover, the mean annual temperature in St. Catherine during the period 2004–2007 ranges from 16.5 to 18.1°C (Dadamouny 2009). Within the mountains that surround St. Catherine there are many wadis subjected to climate change (e.g., Feiran, Zaghra, El-Arbaen). The annual rainfall in W. Feiran was recorded as around 50 mm until the end of the 1970s, but from 1980 until now, the wadi became more arid and annual precipitation decreased (Dadamouny 2009). The cumulative rainfall during extreme events from the summits draining into W. Feiran causes the most severe floods in Sinai (Abd El-Wahab 1995). For some time, the dry valley becomes a mighty river carrying boulders and gravel. In contrast, W. Feiran is the hottest wadi in southern Sinai, especially in June to August (maximum monthly temperature was 42°C in June). In winter, temperatures stay well above zero (7°C, Ramadan 1988). The potential evapotranspiration of St. Catherine was calculated as ca. 1,750 mm yr−1 (Tolba and Gaafer 2003; Cools et al. 2012). Finally, humidity or the amount of water vapor in the air is one of the major factors, since the relative humidity has a great impact on the growth performance of many plant and fungal species. It is also an indicator of the probability of precipitation, dew, or fog, and therefore one can use the relative humidity in weather forecasts. 5.2. Impact of climate change on biological and natural resources Obviously, a lack of precipitation has great difficult impacts on the natural vegetation of South Sinai, where, many plant species disappear or remain with older individuals only (like the trees Acacia tortilis, Crataegus × sinaica and Moringa peregrina) which show low vitality and often lacking recruitment, Dadamouny 2009). Also, this hampers recruitment (Dadamouny et al. 2012). According to Greenwood (1997), the number of species in Sinai deserts is further influenced by the steepness of climatic gradients. To assess the biodiversity response to climate change, Hannah et al. (2002) emphasized the need to apply simulation models operating at a regional scale. Moreover, species respond differently to climate change because of different adaptations to their environment (Erasmus et al. 2002). As a consequence, single-species models with a regional focus are essential (Tews and Jeltsch 2004) to understand the manifold impact of global climate change entirely. This is what we want to achieve by recording and forecasting the population size of Moringa peregrina as one of the rare plant species in Sinai. However, even though recent simulation tools have occasionally been applied to climate-sensitive animal species (Forchhammer et al. 1998; Peterson and Kwak 1999; Saether et al. 2000; Wang et al. 2002; Wichmann et al. 2003), spatial plant population models are extremely rare (Kickert 1999). As climate change may put half of the North American bird species at risk of extinction (Shea 2014), the same

39 Publications Trends of climate with a rapid change in Sinai, Egypt

situation may occur in Sinai. Conservation efforts should be directed to conserve biological resources, and use them in a sustainable way to generate a constant income for local inhabitants. The responses of organisms towards climate change include phenotypic plasticity, genetic adaptation, and dispersal, migration or range shifts (Valladares et al. 2014), and these responses need further studies.

Precautionary measures should include the establishment of institutional coastal and biodiversity monitoring capabilities as recommended by El-Raey (2011), enforcing laws and regulations (El-Raey 2010). Also, consistent coastal zone controlling, raising public awareness of climate risks, and adaptation plans are needed. Based on the first scenario that the sea level may rise around 30 cm, Egypt may lose some parts of its coasts which extend over 3,500 kilometers. With one-third of this distance running along the Mediterranean Sea from Rafah city in the east to Salloum in the west, and two-thirds along the Red Sea and the coasts of the Sinai Peninsula. This risk would induce a severe economic crisis, considering that about 53% of Egypt’s population lives within 100 kilometers of the coasts (EEAA 2005). Also, Egypt has a large number of inland lakes such as Lake Nasser (the largest freshwater one) and Qarun (the saline lake) in Fayyoum. Already a modest rise of sea level will severely affect most of the people along the coasts of the Mediterranean Sea in Egypt, as expected also in China, particularly in the Pearl River Deltas plain containing the city of Guangzhou (Han 1989). Another problem is the water supply for the local inhabitants in North Sinai and the coastal area of the Suez Gulf. In general, Sinai has three indigenous water sources: rainfall, surface water and groundwater (Anonymous 1985; Zahran and Willis 2009). Most of the water comes from the Nile, which is treated at Port Said. But, the city of El-Tor and most of South Sinai depend on groundwater. The coastal area of the Gulf of Aqaba depends on desalinated water. At the same time, the northern strip of Sinai has received the highest amount of rainfall (300 mm/yr, Abdel-Wahab 2003). According to Hammad (1980), the amount southward was (25 mm/yr). With the increase of the sea level, the amount of groundwater is subjected to decline, and this is the minimum impact of climate change. Without an effective program for coastal protection a minimal sea-level rise is likely to affect many coastal cities (Port Said, Damietta, and Alexandria) and those of the northern delta (Kafr El-Sheikh, Tanta, and El-Mahalla El-Kubra). Unfortunately, Egypt is thought of as one of the countries most susceptible to the pronounced adverse impacts of climate change (El-Raey 2010). Recent efforts to understand and mitigate the enormous impacts of climate change involve the development of Egypt’s climate change action plan (CCAP). This also includes Egypt’s national communication on climate change and seeking support from international donor authorities and organizations (EEAA 1999). Moreover, an early warning system for flash floods has been developed for parts of the Sinai Peninsula to understand the response of the wadi to rainfall and observed flash flood events by Cools et al. (2012) and Eissa et al. (2014). A key factor is protective dams in wadis to protect settlements from flooding and mudflows from sediments and rubble supplied by intense physical weathering in the deep gorges. It is reported that scant rainfall on the mountains is one of the main water resources in southern Sinai, which runs over the slopes and collects in narrow-deep wadis forming perpetual streams and rivulets (Zahran and Willis 2009). But, the excess water in the rainy years could percolate and store in rock crevices (Abd El-Ghani and Amer 2002). Outside of the focus of this study are possible impacts on the delta of the Nile River by sea-level rise and saltwater intrusion. Scenarios of sea-level rise by one meter by 2050 as a result of

40 Trends of climate with a rapid change in Sinai, Egypt Publications

global warming are estimated to extinguish about one-third of the Nile Delta (Tsimplis and Shaw 2010; Calafat et al. 2012; IPCC 2013). 5.3. Proposed solutions Egypt has ratified a number of multilateral environmental agreements and already has certain national level environmental plans that intersect with responses that might be required to manage climate variability and long- term climate change. We would like to emphasize the following points: (1) Public awareness and concern for environmental problems are growing, but the complexity of the problems and the uncertainty about many basic data quite often make discussions inconclusive; even indications issued by scientific authorities are sometimes misleading, and the problems are exacerbated by the frequent influence of ideological positions (Wu and Sardo 2010). Therefore, we recommend increasing public awareness in dealing with climate change issues. (2) To keep pace with the requirements for the population growth, agriculture production has to increase by 70% by 2050 as recommended by El-Ramady et al. (2013). The big challenge will be to reconcile this growth with the conservation of the environment, simultaneously reducing the risks associated with increasing emissions of greenhouse gases and pollution with heavy metals (Aune 2012). (3) Egypt is a sunny country. Thus, solar energy has a high potential for use in water desalination, air conditioning, refrigeration and other requirements of life. (4) The increase of wind speed on the coastal line and in South Sinai (El-Tor, , and St. Catherine) is another source of renewable energy (windmills). (5) The dykes and protective measurements on the sea coasts may help to overcome the sea-level increase, but could cause groundwater salination. (6) Establishing a number of desalination stations along with the coasts of the Mediterranean Sea and Red Sea should be considered. (7) Economic development may be fostered by establishing the New Suez Canal. It is a proposed waterway parallel to the existing Suez Canal which is the shortest shipping route between and Asia (164 km long in 1869 and 193.3 km long in 2010). The New Suez Canal is expected to be only 72 km long but requires 35 km of new trenches and 37 km of expansion and deepening of the existing canal. Also, it requires the widening of the Great Bitter Lakes bypasses and Ballah bypass. The new canal between the Red Sea and the Mediterranean Sea may be a measure to accommodate the increase of sea level on both coasts. 5.4. Conclusion Sinai is characterized by an extremely arid climate with pronounced extremes in both temperature and rainfall, sparse vegetation, rare fauna, and precious natural resources. Our study evaluated daily weather records between 1970 and 2014 in six regions in Sinai (El-Arish, Ras Sedr, St. Catherine, Sharm El-Sheikh, El Tor, and Taba). As the climate model is based on many mathematical equations controlled by physical laws, it shows more detailed specifications on the climate of a particular location with lower errors. But, here we simply used the actual trend based on the data set measured by authorized stations, to reveal the previous and the current status of Sinai climate. Based on the regression equations, we found trends of decreasing precipitation combined with increasing temperature (average increase +2.3°C from 1970 to 2014), which will increase the level of aridity. Also, extreme events, like flash floods, seem to become more frequent and unpredictable after long years of drought, carrying a significant amount of total precipitation. These events occur in an unpredictable manner and vary locally as convective rains. These pathways could continue with the same trend to get the forecasted values in the upcoming years (for example, temperature in 2050) or could be interrupted by a new change that could return the weather parameters to the normal pathways (less frequent). Not only our study area, 41 Publications Trends of climate with a rapid change in Sinai, Egypt

but also the whole of Egypt, requires more studies to build an appropriate model that could match with the global pathways or at least differ locally in space and time. Based on the extreme values during recent decades, the trend of temperature in Egypt and the Arabian Peninsula could exceed our forecasting. Also, scarcity of rainfall for many years could negatively affect the growth of wild and medicinal vegetation. During recent decades, we recorded unusually high rainfall (like 205.5 mm in 2002) at El-Arish on the Mediterranean Sea. This is followed by Ras Sedr (94.5 mm in 2006) and El-Tor (82.6 mm in 2002). These amounts are recorded for one to three days of the year, demonstrating another problem: that the high amount of rainfall in Sinai doesn't normally distribute in the months of the winter season. Extreme rainfalls or flash floods could harm important locations in Sinai like the flood of February 1975 at El-Arish drainage basin and the flood of 2010 at St. Catherine. Therefore, it is important to establish protective dams in many areas of the Sinai like the wadis Watir, Zaghra, and El-Arbaeen to protect settlements, and also natural and biological resources. As a result of an expected increase of the sea level, parts of the Egyptian coasts with low elevation above sea level may be lost. This may include Nuweiba area on the Red Sea and many cities on the Mediterranean Sea and the north of the delta. Consequently, a second canal between the Red Sea and the Mediterranean Sea at an appropriate depth could help to accommodate the increase in sea levels, and more studies on this concern are recommended. Care efforts should be directed to conserve the biological and natural resources and to keep pace with climate change.

Acknowledgements The first author was funded by the German Academic Exchange Service (DAAD) in the framework of the German-Egyptian Research Long-term Scholarship program. The authors acknowledge all staff of the Egyptian Meteorological Authority, for helping with data collection. Appreciation extends to Prof. Ahmed Abdel-Aal, Under Secretary of State for Research and Climate, EMA, Egypt for making accessible the reports of the stations located in the study area. We thank the staff members of EMA (Mr. Abdul Rahman Soliman, Mr. Kamal, and Mr. Hamdy Abdel Rahman) for their support during data collection. Thanks to Mrs. Riham Elghandour for her help in computing the data. We wish to thank as well Dr. Martin Unterseher, Institute of Botany and Landscape Ecology, University Greifswald, Germany for reviewing the manuscript. The views expressed in this study are those of the authors and do not reflect the views or policies of EMA.

References Abdel-Aziz, T.M. (2011). Climate change modeling activities undertaken in Egypt. EGM on Climate Change, Beirut, Lebanon, http://css.escwa.org.lb/SDPD/1608/3b.pdf Abdelaziz, R. and Bakr, M. (2012). Inverse modeling of groundwater flow of Delta Wadi El-Arish. Journal of Water Resource and Protection 4, 7, doi:10.4236/jwarp.2012.47050 Abd El-Ghani, M.M. and Amer, W.M. (2002).Soil–vegetation relationships in a coastal desert plain of southern Sinai, Egypt. Journal of Arid Environments 55, 607–628. Abd El-Wahab, R.H. (1995). Reproduction ecology of wild trees and shrubs in southern Sinai, Egypt, M.Sc. diss. Botany Department, Faculty of Science, Suez Canal University, Ismailia, Egypt, 38pp. Abd El-Wahab, R.H. ( 2003 ).Ecological Evaluation of Soil Quality in South Sinai, Egypt, Ph.D. Thesis, Botany Department, Faculty of Science, Suez Canal University, Ismailia, Egypt. 247 pp. Ahmed, M.T, Nagi, I., Farag, M., Loutfi, N., Osman, M.A., Mandour, N.S., Mohamoud K. and Loutfi, N. (2014). Vulnerability of Ras Sudr, Egypt to climate change, livelihood index, an approach to assess risks and develop future adaptation strategy. Journal of Water and Climate Change, Doi:10.2166/wcc. (2014). 006 Anonymous, E. (1985). Sinai Development Study, Phase I. Final report Volume V. Water supplies and costs. Dames and 42 Trends of climate with a rapid change in Sinai, Egypt Publications

Moore, Chapter 2. Water Resources Assessment, 90 pp. Arrhenius, S. (1896). On the influence of carbonic acid in the air upon the temperature of the ground. London, Edinburgh, and Dublin Philosophical Magazine and Journal of Science 41 (5), 237–275. Arrhenius, S. (1908). Das Werden der Welten. Academic Publishing House, Leipzig. Aune, J.B. (2012). Conventional, organic and conservation agriculture: Production and environmental impacts. E. Lichtfouse (Ed.). Agroecology Reviews 8. Doi:10.1007/987-94-007-(1905-7_7. Ayyad, M.A. and Ghabbour, S.I. (1986). Hot deserts of Egypt and Sudan, In: Ecosystems of the World, 12B, Hot Desert and Arid (M. Evenari, I. Noy-Meir and D.W. Goodall, Eds.). Elsevier, Amsterdam, pp. 149–202. Batanouny, K.H. (1981). Eco-physiological studies on desert plants. X-Contribution to the autoecology of the desert chasmophyte, Stachys aegyptiaca Pers. Oecologia 50, 422–426. Bates, B.B., Kundzewicz, Z.W., Wu, S. and Palutikof, J.P. (2008). Climate Change and Water. Technical Paper of the Intergovernmental Panel on Climate Change, Geneva. Calafat, F.M., Chambers, D.P. and Tsimplis, M.N. (2012). Mechanisms of decadal sea-level variability in the eastern North Atlantic and the Mediterranean Sea. Journal of Geophysical Research: Oceans 117, C09022. Chanton, J.P., Bauer, J.E., Glaser, P.A., Siegel, D.I., Kelley, C.A., Tyler, S.C., Romanowich, E.H. and Lazrus, A. (1995). Radiocarbon evidence for the substrates supporting methane formation within northern Minnesota peatlands. Geochimica et Cosmochimica Acta 50 (17), 3663–3668. Cools, J., Vanderkimpen, P., El Afandi, G. Abdelkhalek, A. Fockedey, S. El Sammany , M., Abdallah, G., El Bihery, M., Bauwens, W. and Huygens, M. (2012). An early warning system for flash floods in hyper-arid Egypt. Natural Hazards and Earth System Sciences 12, 443–457, www.nat-hazards-earth-syst-sci.net/12/443/2012/ Doi:10.5194/nhess-12- 443–2012. Comsan, M. (2011). Energy sector and climate change. International Conference on Energy Systems and Technologies (ICEST 2011) 11-14 March (2011). , Cairo, Egypt. Dadamouny, M.A. (2009). Population Ecology of Moringa peregrina growing in southern Sinai, Egypt. M.Sc. Thesis, Environmental Sciences, Faculty of Science, Suez Canal University, Egypt, 205 pp. Dadamouny, M.A., Zaghloul, M.S. and Moustafa, A.A. (2012). Impact of improved soil properties on establishment of Moringa peregrina seedlings and a trial to decrease its mortality rate. Egyptian Journal of Botany NIDOC. 07/2012. Dames and Moore (1982). Sinai Development Study. Phase I Final Report, Advisory Committee for Reconstruction, Ministry of Development, Egypt in association with industrial development programs, SA. Danin, A. (1986). Flora and vegetation of Sinai. Proceedings of the Royal Society of Edinburgh 89 (B), 159–168. Duraiappah, A.K. (2006). Millennium Ecosystem Assessment: Ecosystems and Human-well Being-biodiversity Synthesis. World Resources Institute, Washington, DC. ISBN 1-56973-588-3. Egyptian Environmental Affairs Agency (EEAA) (1999). The Arab Republic of Egypt: Initial National Communication on Climate Change. Report for the United Nations Framework Convention on Climate Change UNFCCC. Egyptian Environmental Affairs Agency (EEAA) (2011). St. Katherine Protected Area of Egypt. Final Report of (2011). EEAA and UNEP (1995). UNEP Greenhouse Gas Abatement Costing Studies: Case Study On Egypt, Report (1995). Eissa, M.A., Thomas, J.M., Hershey, R.L., Dawoud, M.I., Pohll, G. and , K.A. (2014). Geochemical and isotopic evolution of groundwater in the Wadi Watir watershed, Sinai Peninsula, Egypt. Environmental Earth Sciences 71, 1855–1869. Doi:10.1007/s12665-013-2588-4 El-Bihary, M.A. and Lachmar, T.E. (1994). Groundwater quality degradation as a result of overpumping in the delta Wadi El-Arish area, Sinai Peninsula, Egypt. Journal of Environmental Geology, 24, 293–305. El-Raey, M. (2010). Impacts and implications of climate change for the coastal zones of Egypt. In: Coastal Zones and Climate Change (D. Michel and A. Pandya, Eds.). Stimson, Washington, pp. 31–50. El-Raey, M. (2011). Mapping areas affected by sea-level rise due to climate change in the Nile Delta until 2100. In: Coping with Global Environmental Change, Disasters and Security (H. Günter, U.O. Spring, C. Mesjasz, J. Grin, P. Kameri- Mbote, B. Chourou, P. Dunay and P. Birkmann, eds.). Springer-Verlag, Germany, pp. 773–788. El-Ramady, H.R., El-Marsafawy, S.M. and Lewis, L.N. (2013). Sustainable agriculture and climate changes in Egypt, E. Lichtfouse (Ed.). Sustainable Agriculture Reviews 12, 41. doi:10.1007/978-94-007-5961-9_2, Springer (2013). Emanuel, K.A. (1987). The dependence of hurricane intensity on climate. Nature 326, 483–485. Erasmus, B.N., Van Jaarsveld, A.S., Chown, S.L., Kshatriya, M. and Wessels, K.J. (2002). Vulnerability of South African animal taxa to climate change. Global Change Biology 8, 679–693. Field, R., Hawkins, B.A., Cornell, H.V., Currie, D.J., Diniz-Filho, J.F., Guégan, J., Kaufman, D.M., Kerr, J.T., Mittelbach, G.G., Oberdorff, T., O’Brien, E.M. and Turner, J.G. (2009). Spatial species-richness gradients across scales: a meta- analysis. Journal of Biogeography 36 (1), 132–147. Doi:10.1111/j.1365-2699.2008.01963.x Food and Agriculture Organization (FAO) (2008). Climate change and food security: a framework document. 43 Publications Trends of climate with a rapid change in Sinai, Egypt

Interdepartmental Working Group on Climate Change, Rome, (2008). Forchhammer, M.C., Stenseth, N.C., Post, E. and Langvatn, R. (1998). Population dynamics of Norwegian red deer: density-dependence and climatic variation. Proceedings of the Royal Society of London Series B-Biological Sciences 265, 341–350. Gillad, D. (1975). The flood in Wadi El-Arish, hydrological survey. Hydrological Survey of Israel, unpublished interim report. Golitsyn, G.S. (1989). Climate changes and related issues. Presentation 24 August (1989). at Sundance, Utah Symposium on Global Climate Change, Institute of Atmospheric Physics, Moscow, USSR. González-Aparicio, I., Baklanov, A., Hidalgo, J., Korsholm, U., Nuterman, R. and Mahura, A. (2014). Impact of city expansion and increased heat fluxes scenarios on the Urban Boundary Layer of Bilbao using Enviro-HIRLAM. Urban Climate Journal, SI: Cities and Megacities. doi:10.1016/j.uclim.2014.07.010 Google Earth (2014). Google Earth Outreach Tutorials. Geological Survey, NASA, USA. http://earth.google.com/outreach/tutorials.html Gorham, E. (1991). Northern Peatlands: role in the carbon cycle and probable responses to climatic warming. Ecological Applications 1 (2), 182–195. Greenwood, N. (1997). The Sinai: A Physical Geography, 1st ed. University of Texas Press, 2nd ed. (2010). , 247 pp. Hammad, F.A. (1980). Geomorphological and hydrogeological aspects of Sinai Peninsula, A.R.E. Annals of the Geographical Survey of Egypt X, 807–817. Han, M. (1989). Global warming induced a sea-level rise in China: Response and strategies. Presentation to World Conference on Preparing for Climate Change, December 19, Cairo, Egypt. Hannah, L., Midgley, G.F., Lovejoy T., Bond, W.J., Bush, M., Lovett, J.C., Scott, D. and Woodward, F.I. (2002). ).Conservation of biodiversity in a changing climate. Conservation Biology 16, 264–268. Hansen, J., Rind, D., Del Genio, A., Lacis, A., Lebedeff, S., Prather, M., Ruedy, R. and Karl, T. (1989). Regional greenhouse climate effects. In: Coping With Climate Change (J.C. Topping, Jr., ed.). Climate Institute, Washington, DC, USA. Henderson-Sellers, A. and Blong, R. (1989). The Greenhouse Effect: Living in a Warmer Australia, New South Wales University Press, Kensington, NSW, Australia. Hoath, R. (2009). A Field Guide to the Mammals of Egypt. American University in Cairo Press, (2009). , Nature, 320 pp. Howard, G., Charles, K., Pond, K., Brookshaw, A., Hossain, R. and Bartram, J. (2010). Securing 2020 vision for 2030: climate change and ensuring resilience in water and sanitation services. Journal of Water and Climate Change 2–16. Doi: 10.2166/wcc.2010.205 Howard, J, Babij E, Griffis R, Helmuth B, Himes-Cornell A, Niemier P, Orbach M, Petes L, Allen S, Auad G, Auer C, Beard R, Boatman M, Bond N, Boyer T, Brown D, Clay P, Crane K, Cross S, Dalton M, Diamond J, Diaz R, Dortch Q, Duffy E, Fauquier D, Fisher W, Graham M, Halpern B, Hansen L, Hayum B, Herrick S, Hollowed A, Hutchins D, Jewett E, Jin D, Knowlton N, Kotowicz D, Kristiansen T, Little P, Lopez C, Loring P, Lumpkin R, Mace A, Mengerink K, Morrison JR, Murray J, Norman K, O'Donnell J, Overland J, Parsons R, Pettigrew N, Pfeiffer L, Pidgeon E, Plummer M, Polovina J, Quintrell J, Rowles T, Runge J, Rust M, Sanford E, Send U, Singer M, Speir C, Stanitski D, Thornber C, Wilson C, Xue Y. 2013 Oceans and marine resources in a changing climate. Oceanography and Marine Biology: An Annual Review 51, 71–192. Ingole, S.P. and Kakde, A.U. 2013 Global warming and climate change: impact on biodiversity. International Journal of Scientific Research 2 (5), 288–290. Intergovernmental Panel on Climate Change (IPCC) 1990 Potential impacts of climate change. Report of Working Group 2, Intergovernmental Panel on Climate Change, 1–1 to 2. World Meteorological Organization (WMO)/United Nations Environment Programme (UNEP), Geneva. Intergovernmental Panel on Climate Change (IPCC) 2013 Climate Change 2013: The Physical Science Basis, Working Group I Contribution to the Fifth Assessment Report of the Intergovernmental Panel on Climate Change, Cambridge University Press, 1535 pp. Intergovernmental Panel on Climate Change (IPCC) 2014 Climate Change 2014: Mitigation of Climate Change. The Fifth Assessment Report of the Intergovernmental Panel on Climate Change, Cambridge University Press, 1435 pp.

International Energy Agency IEA 2013 CO2 Emission from Fuel Consumption Highlights. IEA Publications, France, 158 p. Issar, A. and Gilad, D. 1982 Ground water flow systems in the arid crystalline province of southern Sinai. Journal of Hydrogeology Science 27 (3), 309–325. Jay, D.A. and Naik, P.K. 2002 Separating human and climate impacts on Columbia River hydrology and sediment transport. In: Southwest Washington Coastal Erosion Workshop Report 2000 (G. Gelfenbaum and G.M. Kaminsky, 44 Trends of climate with a rapid change in Sinai, Egypt Publications

eds.). U.S. Geological Survey Open-File Report 02-229, pp. 38–48. Jay, D.A. and Naik, P.K. 2011 Distinguishing human and climate influences on hydrological disturbance processes in the Columbia River, USA. Hydrological Sciences Journal 56 (7), 1186–1209. Kassas, M. and Girgis, W.A. 1970 Habitat and plant communities in the Egyptian desert. VII. Geographical facies of plant communities. Journal of Ecology 58, 335–350. Kerr, R.A. 1990 Global warming continues in 1989. Science February 2, 1990, p. 521. Kickert, R.N., Tonella, G., Simonov, A. and Krupa, S.V. 1999 Predictive modeling of effects under global change. Environmental Pollution 100, 87–132. Klein, M. 2000 The formation and disappearance of a delta at the El-Arish river mouth. The Hydrology-Geomorphology Interface: Rainfall, Floods, Sedimentation, Land Use (Proceedings of the Conference, May 1999). IAHS Publication 261, 2000. Lal, R. and Stewart, B.A. 2012 Sustainable management of soil resources and food security. In: World Soil Resources and Food Security (R. Lal and B.A. Stewart, Eds.). CRC Press, USA, 574 pp. Lappalainen, E. 1996 Global Peat Resources. International Peat Society, Jyskä, Finland. Lichtfouse, E. 2009 Climate change, society issues and sustainable agriculture. In: Climate Change, Intercropping, Pest Control and Beneficial Microorganisms (E. Lichtfouse, Ed.). Sustainable Agriculture Reviews, Vol. 2. Springer, pp. 1–7. Doi: 10.1007/978-90-481-2716-0_1 Lin, X. 1999 Flash Floods in Arid and Semi-arid Zones. UNESCO, Paris, 65 pp. Meinshausen, M., Smith, S.J., Calvin, K., Daniel, J.S., Kainuma, M.L., Lamarque, J-F., Matsumoto, K., Montzka, S.A., Raper, S.C.B., Riahi, K. Thomson, A., Velders, G.J.M. and van Vuuren, D.P.P. 2011 The RCP greenhouse gas concentrations and their extensions from 1765 to 2300. Climatic Change 109 (1–2), 213–241, doi:10.1007/s10584- 011-0156-z Migahid, A.M., El-Shafei, A., Abd El-Rahman, A.A. and Hammouda, M.A. 1959 Ecological observations in Western and Southern Sinai. Bulletin Social Geography d'Egypte 32, 166–206. MINITAB 2014 Meet MINITAB 17 for Windows, a concise guide to getting started with Minitab software [online http://www.minitab.com/en-us/products/minitab/]. Moustafa, A.A. 1990 Environmental Gradients and on Sinai Mountains. Ph.D. Thesis, Botany Department, Faculty of Science, Suez Canal University, Ismailia, Egypt, 115 pp. Murday, M. 1989 Presentation to Conference on Implications of Climate Change for Africa. Howard University, Washington, DC, USA, 4. May 1989. Naik, P.K. and Jay, D.A. 2011 Distinguishing human and climate influences on the Columbia River: Changes in mean flow and sediment transport. Journal of Hydrology 404 (3–4), 259–277. Nir, Y. 1982 Offshore artificial structures and their influence on the Sinai Mediterranean beaches. In: Proceedings of. Eighteenth Coastal Engineering Conference, ASCE, Capetown, South Africa, Nov. 14–19, pp. 1837–1856. Nir, Y. 1989 The Mediterranean Sea coast in Sinai: sedimentological aspects. Geological Survey 39–88. Nordstokke D.W. and Zumbo B.D. 2010 A new nonparametric Levene's test for equal variances. Psicologica 31, 401–430. Nordstokke, D.W., Zumbo, B.D., Cairns, S.L. and Saklofske, D.H. 2011 The operating characteristics of the nonparametric Levene’s test for equal variances with assessment and evaluation data. Practical Assessment, Research and Evaluation 1 (5). Peel, M.C. Finlayson, B.L. and McMahon, T.A. 2007. Updated world map of the Köppen–Geiger climate classification. Hydrology and Earth System Sciences 11, 1633–1644. Peterson, J.T. and Kwak, T.J. 1999 Modeling the effects of land use and climate change on riverine smallmouth bass. Ecological Applications 9, 1391–1404. Pressey, R.L., Cabeza, M., Watts, M.E., Cowling, R.M. and Wilson, K.A. 2007 Conservation planning in a changing world. Trends in Ecology and Evolution (Amst.) 22 (11), 583–92. doi:10.1016/j.tree.2007.10.001. PMID 17981360. Ramadan, A.A. 1988 Ecological Studies in Wadi Feiran, its Tributaries and the Adjacent Mountains. Ph.D. Thesis, Faculty of Science, Suez Canal University, Egypt, 307 pp. Saether, B.E., Tufto, J., Engen, S., Jerstad, K., Rostad, O.W. and Skatan, J.E. 2000 Population dynamical consequences of climate change for a small temperate songbird. Science 287, 854–856. Said, R. 1990 The Geology of Egypt. Elsevier, Amsterdam, 734 pp. Sala O.E., Chapin, F.S. and Armesto, J.J. 2000 Global biodiversity scenarios for the year 2100. Science 287 (5459), 1770– 1774. doi:10.1126/science.287.5459.1770. PMID 10710299. Schneider, S.H. 1989 The greenhouse effect: reality or media event. In: Coping With Climate Change (J.C. Topping, Jr., Ed.). Climate Institute, Washington, DC, USA. Shea, R.H. 2014 Climate change could threaten half of North American birds by the end of the century. National Audubon 45 Publications Trends of climate with a rapid change in Sinai, Egypt

Society, USA. Sept., 8. 2014. http://www.audubon.org/ Shokry, M. and Ammar, A. 2009 Assessment of present status and future needs of four Coral Reef sites along the Gulf of Aqaba, Egypt. The Open Environmental Pollution and Toxicology 1, 34–42. Smith, S.E., El-Shamy, I. and Abd-El Monsef, H. 1997 Locating regions of high probability for groundwater in the Wadi EI-Arish Basin, Sinai, Egypt. Journal of African Earth Sciences 25 (2), 253–262. SPSS. IBM Corp; Armonk, NY: 2013 IBM SPSS Statistics for Windows, Version 22.0. IBM Corp. Released. Stern, N. 2006 The Economics of Climate Change. Cambridge University Press, Cambridge. Tews, T. and Jeltsch, F. 2004 Modelling the impact of climate change on woody plant population dynamics in South African savannah. BMC Ecology 4–17. doi:10.1186/1472-6785-4-17. Tickell, S.C. 1989 Environmental Refugees. National Environment Research Council Annual Lecture, the Royal Society, London, UK, 5 June 1989. Titus, J.G. 1988 Cause and effects of sea-level rise. In: North American Conference on Preparing for Climate Change, a cooperative approach, October 27-29, 1987, Washington, D.C. Climate Institute, Inc., Rockville, Maryland, USA. Tolba, A. F. and Gaafer, K. 2003 On estimation of potential evapotranspiration in Egypt. Egyptian Meteorological Authority, Meteorological Research Bulletin, 1687–1014. Tsimplis, M. N. and Shaw, A.G.P. 2010 Seasonal sea-level extremes in the Mediterranean Sea and at the Atlantic European coasts. Natural Hazards and Earth System Sciences 10, 1457–1475. UNEP 1992 World Atlas of Desertification. Edward Arnold, London. Valladares, F., Matesanz, S., Guilhaumon, F., Araújo, M.B, Balaguer, L., Benito-Garzón, M., Cornwell, W., Gianoli, E., Kleunen, M., Naya, D.E., Nicotra, A.B., Poorter, H. and Zavala, M.A. 2014 The effects of phenotypic plasticity and local adaptation on forecasts of species range shifts under climate change. Journal of Ecology Letters 17 (11), 1351– 1364. Doi: 10.1111/ele.12348 Wang, G.M., Hobbs, N.T., Singer, F.J., Ojima, D.S. and Lubow, BC. 2002 Impacts of climate changes on elk population dynamics in Rocky Mountain National Park, Colorado, USA. Climatic Change 54, 205–223. Weart S.R. 2008 The Discovery of Global Warming. Harvard University Press. 240 pp. Wichmann, M.C., Jeltsch, F., Dean, W.R.J, Moloney, K.A. and Wissel, C. 2003 Implications of climate change for the persistence of raptors in arid savanna. Oikos 102, 186–202. Wu, J. and Sardo, V. 2010 Sustainable versus organic agriculture. In: Sociology, Organic Farming, Climate Change and Soil Science (E. Lichtfouse, Ed.). Sustainable Agriculture Reviews, 3, Springer. doi: 10.1007/978-90-481-3333-8-11 Xu, J., Shrestha, A.B., Vaidya, R., Eriksson, M. and Hewitt, K. 2007 The Melting Himalayas-Regional Challenges and Local Impacts of Climate Change on Mountain Ecosystems and Livelihoods. ICIMOD Technical Paper. International Centre for Integrated Mountain Development (ICIMOD), Kathmandu, Nepal. Zahran, M.A. and Willis, A.J. 2009 The Vegetation of Egypt, 2nd Ed. Chapman and Hall, London, 437 pp. Zohary, M. 1973 Geobotanical foundations of the . Geobot. Selecta, 3, Gustav Fischer Verlag (2), Stüttgart, 739. First received 18 January 2015; accepted in revised form 11 September 2015. Available online 16 October 2015

Appendix I Appendix 1. Fig. S1. (a-c) Impact of drought as a result of scarce rainfalls on Sinai through the last 20 years, and (d) St. Catherine in Apr. 2011 after flash flood that occurred in 2010.

Appendix 1. Fig. S2. (a) Satellite map showing circumscriptions of the water drainage area of the Wadi El-Arish in Sinai. (b) Dams or walls established within wadi beds to mitigate floods threatening rare populations of plants and Bedouin houses in W. Zaghra, South Sinai, (c) Map of Wadi Watir in South Sinai, and (d-g) Maps taken from Google Earth, showing W. Watir as one of the wadis most affected by floods.

Appendix 1. Fig. S3. Landscapes of Sinai affected by severe drought or unpredictable rainfalls. (a) Snowfall at St. Catherine in 2013, (b) Snow covering trees, camels, and mountains of St. Catherine, and (c-f) A flash flood at the Sinai region in 2010.

Appendix II Table S1. Annual data for seven climate parameters recorded at six weather stations in the Sinai Peninsula during the last twenty years (1995–2014). Climate parameters are maximum, mean and average temperature (ºC), total rainfall (mm), and means for relative humidity (%), wind speed (km/hr), and air pressure at sea level (hPa).

46

Trends of climate with a rapid change in Sinai, Egypt

Mohamed A. Dadamouny and Martin Schnittler Ernst-Moritz-Arndt University of Greifswald, Institute of Botany and Landscape Ecology, Soldmannstr. 15, D-17487 Greifswald, Fax +49 03834 420 4114 *Email: [email protected]

Appendix I

Fig. S1. (a-c) Impact of drought as a result of scarce rainfalls on Sinai through the last 20 years, and (d) St. Catherine in Apr. 2011 after flash flood that occurred in 2010.

47 Appendix 01 Trends of climate with a rapid change in Sinai, Egypt

Fig. S2. (a) Satellite map showing circumscriptions of the water drainage area of the Wadi El-Arish in Sinai. (b) Dams or walls established within wadi beds to mitigate floods threatening rare populations of plants and Bedouin houses in W. Zaghra, South Sinai, (c) Map of Wadi Watir in South Sinai, and (d-g) Maps taken from Google Earth, showing W. Watir as one of the wadis most affected by floods.

48 Trends of climate with a rapid change in Sinai, Egypt Appendix 01

Fig. S3. Landscapes of Sinai affected by severe drought or unpredictable rainfalls. (a) Snowfall at St. Catherine in 2013, (b) Snow covering trees, camels, and mountains of St. Catherine, and (c-f) A flash flood at the Sinai region in 2010.

49 50 Appendix II

Appendix II

Table S1. Annual data for seven climate parameters recorded at six weather stations in the Sinai Peninsula during the last twenty years (1995–2014). Climate parameters are maximum, mean and average temperature (ºC), total rainfall (mm), and means for relative humidity (%), wind speed (km/hr), and air pressure at sea level (hPa). Means are based on daily records.

El-Arish 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 Max. Temp. 41.0 44.0 42.0 43.0 42.0 39.0 40.0 42.0 45.0 43.0 39.4 42.0 45.0 42.0 39.0 41.0 40.0 39.0 45.0 44.0 Mean Temp. 20.2 20.1 18.4 20.8 20.5 19.7 20.7 21.1 20.3 20.2 21.4 23.2 20.4 20.6 20.7 21.7 22.8 20.3 20.4 20.7 Min. Temp. 0.0 1.0 3.0 0.0 2.0 1.0 1.0 1.0 2.0 0.0 2.0 2.0 3.0 -3.0 2.0 3.0 3.0 2.0 4.0 4.0 T. Rainfall 9.2 46.7 116. 143. 13.9 107. 79.7 205. 75.9 139. 33 150. 121. 142. 33.5 119. 62.0 44.6 47.7 48.2 Humidity 65.9 66.8 68.1 67.0 68.6 70.6 68.2 67.9 66.6 67.6 66.5 76.0 70.2 66.7 66.7 54.7 65.1 69.7 69.0 73.3 Wind Speed 11.0 7.4 8.2 7.4 6.1 6.6 7.7 7.1 8.5 7.9 8.8 7.6 6.8 7.0 9.4 8.8 7.6 7.7 8.3 5.4 Air pressure 1015 1012 1015 1013 1013 1014 1013 1013 1013 1014 1014 1016 1012 1013 1012 842. 1010 1012 1013 1013 Ras Sedr 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 Max. Temp. 38.0 39.0 39.0 40.0 41.0 43.0 41.0 45.0 41.0 41.0 44.0 43.0 44.0 42.0 43.0 43.0 41.0 41.0 42.0 42.0 Mean Temp. 22.6 21.4 21.7 21.4 21.8 22.3 21.5 23.1 24.1 22.8 22.5 22.8 23.2 23.1 23.4 24.4 22.6 23.4 23.2 25.4 Min. Temp. 2.0 3.0 1.0 5.0 3.0 4.0 6.0 5.0 3.0 8.0 5.0 5.0 2.0 2.0 6.0 4.0 1.0 4.0 6.0 7.0 T. Rainfall 37.1 23.4 28.5 21.7 24 11.4 6.61 14.2 11.1 31.7 11.2 94.5 2.79 11.4 1.78 6.4 10.7 6.9 1.0 16.8 Humidity 57.7 56.2 54.8 55.1 56.8 55.0 52.6 53.3 52.0 52.4 52.8 52.7 51.8 48.8 48.8 50.2 50.1 49.6 49.2 42.7 Wind Speed 12.8 12.5 12.1 11.6 12.6 13.4 11.1 11.2 10.3 10.0 10.7 8.0 12.0 12.1 12.0 12.9 11.5 11.4 11.8 9.7 Air pressure 1015 1013 1014 1014 1015 1013 1013 1013 1012 1012 1013 1012 1012 1012 1011 1010 1011 1011 1011 1011 St. Catherine 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 Max. Temp. 32.0 29.0 28.0 26.0 30.0 26.0 31.0 35.0 36.0 34.0 38.0 35.0 39.0 38.0 43.0 42.0 39.0 37.0 37.0 35.0 Mean Temp. 16.4 18.1 17.8 16.5 15.7 16.6 18.5 17.5 19.0 17.5 18.0 16.5 20.7 18.1 17.9 20.1 16.9 20.6 17.9 17.7 Min. Temp. 1.0 1.5 0.0 3.0 1.4 1.0 0.0 0.0 2.0 0.0 -1.0 -2.0 1.0 -6.0 -3.0 -1.0 1.0 1.0 -3.0 0.0 T. Rainfall 22 25 38 14.6 nd nd nd nd nd 27.0 25.6 8.6 17.4 4.0 7.6 44.1 1.2 20.1 16.6 4.5 Humidity 31.0 30.9 32.4 25.1 22.3 25.2 27.3 37.5 28.4 39.4 40.5 41.6 20.7 36.7 40.2 38.0 42.9 34.5 40.6 41.0 Wind Speed 12.4 11.3 9.8 9.3 8.6 11.4 10.1 8.7 9.8 8.6 8.3 7.9 8.4 7.8 8.6 9.2 6.9 11.5 10.7 10.4 Air pressure 1021 1019 1020 1021 1018 1020 1020 1019 1021 1018 1020 1016 994. 1020 1020 1020 1019 934. 1020 1020 (Continued )

Trends of climate with a rapid change in Sinai, Egypt

Trends of climate with a rapid change in Sinai, Egypt

Table 2. Continued

Sharm El-Sheikh 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 Max. Temp. 39.0 40.0 43.0 43.0 42.0 44.0 42.0 45.0 42.0 43.0 42.0 43.0 43.0 41.0 41.0 46.0 43.0 43.0 46.0 45.0 Mean Temp. 24.5 25.7 25.2 26.0 26.1 25.5 26.5 26.3 26.1 26.8 25.9 25.6 25.9 26.2 26.1 27.6 25.8 26.7 27.4 28.1 Min. Temp. 7.0 3.0 6.0 12.0 12.0 5.0 10.0 9.0 4.0 5.0 10.0 8.0 10.0 10.0 11.0 12.0 8.0 9.0 9.0 5.0 T. Rainfall nd 17.0 0.4 nd nd 9.0 6.5 8.1 12.1 18.3 5.4 6.0 0.8 0.3 0.0 34.4 12.3 0.0 0.6 0 Humidity 44.8 41.7 39.5 40.2 40.6 39.8 40.6 37.4 42.3 36.6 37.8 36.5 37.2 34.5 35.1 36.8 38.5 35.6 34.5 41.8 Wind Speed 21.5 22.1 20.8 18.6 21.3 19.3 18.7 19.6 18.5 19.4 19.3 18.9 17.8 18.4 16.0 17.3 16.6 14.7 15.3 18.7 Air pressure 1015 1013 1013 1014 1013 1012 1012 1012 1013 1011 1011 1011 1010 1010 1010 1010 1010 1010 1010 1009 El-Tor 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 Max. Temp. 45.0 40.0 41.0 39.0 42.0 41.0 42.0 43.0 43.0 42.0 42.0 44.0 43.0 42.0 41.0 45.0 41.0 43.0 43.0 42.0 Mean Temp. 22.9 23.2 22.9 23.7 23.1 22.0 23.3 23.6 23.3 22.9 23.8 27.0 27.6 28.8 25.8 25.8 23.5 22.3 23.7 25.1 Min. Temp. 7.0 4.0 5.0 7.0 6.0 2.0 4.0 5.0 1.0 2.0 3.0 6.0 4.0 6.0 8.0 6.0 4.0 10.0 5.0 8.0 T. Rainfall 29.7 13.5 1.0 3.1 8.3 11.4 11.9 82.6 3.1 5.4 2.0 0.8 4.9 2.5 0.0 17.4 18.6 7.5 5.3 18.6 Humidity 54.3 55.4 52.5 54.8 57.7 58.2 56.7 56.5 55.0 55.8 56.3 54.9 54.5 54.5 58.6 56.5 52.8 56.7 44.4 43.8 Wind Speed 24.8 24.5 22.4 22.9 23.9 24.2 25.7 25.7 24.8 25.1 25.0 24.4 23.6 23.5 21.7 22.5 24.1 25.7 23.1 23.0 Air pressure 1010 1009 1011 1010 1010 1010 1010 1011 1010 1011 1011 1011 1011 1011 1009 1009 1011 1010 1010 1010 Taba 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 Max. Temp. 40.0 45.0 42.0 40.0 41.0 41.0 41.0 44.0 40.0 41.0 40.0 39.0 41.0 43.0 39.0 43.0 38.0 41.0 40.0 42.0 Mean Temp. 21.0 21.1 20.9 20.6 20.8 23.3 23.1 23.1 23.3 18.9 23.0 23.5 22.5 18.9 20.4 23.5 17.7 23.0 23.4 23.6 Min. Temp. 3.0 0.0 5.0 3.0 4.0 4.0 5.0 4.0 4.0 0.0 0.0 0.0 0.0 0.0 -1.0 4.0 2.0 2.0 6.0 1.0 T. Rainfall 13.3 15.2 3.3 8.6 6.9 11.3 13.4 9.8 4.4 10.2 2.9 6.3 12.0 13.5 2.3 61.5 2.1 5.8 23.2 23.9 Humidity 62.0 61.8 50.8 59.0 57.8 51.6 46.8 45.7 57.4 51.4 50.2 65.3 49.7 43.5 43.8 43.7 46.4 43.0 42.0 53.9 Wind Speed 21.6 19.1 20.2 19.3 18.4 16.5 18.9 16.4 15.4 16.7 15.5 16.6 15.0 15.5 14.7 10.8 13.0 10.1 9.9 17.6 Air pressure 1021 1020 1018 1020 1020 1018 1017 1019 1019 1018 1017 1020 1017 1017 1016 1015 1015 1015 1015 1017

51 Appendix II

3.2 Population performance of Moringa peregrina (Forssk.) Fiori (Moringaceae) at Sinai Peninsula, Egypt in the last decades: Consequences for its conservation

Mohamed A. Dadamouny, Martin Unterseher, Peter König, and Martin Schnittler Ernst-Moritz-Arndt University of Greifswald, Institute of Botany and Landscape Ecology, Soldmannstr. 15, D-17487 Greifswald, Fax +49 03834 420 4114 *Email: [email protected]

Abstract We investigated effects of recent climate change and human impacts on the threat situation of Moringa peregrina, a drought-adapted tree distributed at Sinai, Arabian Peninsula, and Horn of Africa. Four populations were surveyed at Sinai and evaluated for its performance and fitness parameters (2006–2015). Total populations declined from 448 (initial census 2000) to 423 (2006) then 390 trees (2015). Recruitment was naturally very sporadic, as indicated by size and age structures of the populations. Except for 12 saplings established after strong rainfalls in 2010 and already dead in 2015, no trees below 5 cm diameter at breast height were recorded, which points to the lacking of recruitment, at least during the last three decades. Multivariate statistics in response to nine environmental variables (elevation, inclination, exposition, landform, annual precipitation, temperature, browsing, cutting, and GPS coordinates) indicated that once established, Moringa trees are relatively resistant to environmental threats. However, tree performance was significantly better in wadis receiving higher rainfalls. Growth sites where the trees showed the highest performance (wadi beds) are also those most threatened by flash floods and suffering from the highest grazing intensity. Seed production occurs only sporadically, ca. 29% of all trees did not produce seeds in both survey years. Seed set was better on south- facing slopes, where trees are much more susceptible to drought. Considering the trend towards a drier and hotter climate at Sinai, we recommend to include the species on the IUCN Red List, provide an updated threat assessment and recommend actions and suitable locations for facilitated recruitment. Keywords Climate trends, impact assessment, IUCN Red List, Moringa, population dynamics, sustainability 1. Introduction 1.1. Importance of Moringa peregrina Moringa is the sole genus of the family Moringaceae. Its thirteen species (Al Kahtani and Abou-Arab 1993; Olson 2002) are distributed in the dry tropics, represented by bottle trees, sarcorhizal trees, slender trees (e.g. M. peregrina, Olson and Carlquist 2001; Olson 2002), and tuberous shrubs. M. peregrina (Forssk.) Fori is a medium sized tree 3-15 m tall (El-Hadidi et al. 1991; Boulos 1999, Dadamouny 2009, see Fig. S1 and Table S2 for descriptive data). M. peregrina is a highly nutritious plant that produces economically valuable oil, making up 54% of the biomass of dry pods (Somali et al. 1984; Al-Kahtani and Abou-Arab, 1993) with antimicrobial activity (Lalas et al., 2012) and in treatment of Herpes simplex virus 52

Population performance and conservation of Moringa peregrina Publications

(Soltan and Zaki, 2009). It is used as well as a water purifier and animal fodder (Abd El-Wahab, 1995; Dadamouny, 2009). Its roots, leaves, flowers and seeds are used as a medication for tumors and cancer (Hartwell, 1971; El-Alfy et al., 2011). Its pods act as a de-wormer and treat liver and spleen problems and pains of the joints (Dadamouny, 2009). Due to the high content of its proteins, carbohydrates, vitamins (A and C), minerals (Ca, P, Fe, and Cu), and fibres of pods, it can treat malnutrition and diarrhea (Asghari et al. 2015). Its leaves are also a source of minerals, vitamins, and proteins and used as anti-oxidant (Nawash and Horani, 2011; Dehshahri et al., 2012). It is used as well against headaches and to stop bleeding. The high oleic acid content of its seeds has a role in lowering blood lipid, systolic blood pressure and serum cholesterol (Al-Dabbas et al., 2010). It has anti-bacterial and anti-inflammatory effects when it applied to wounds or insect bites (Koheil et al., 2011). Moringa pods with its seeds are very palatable for livestock (Owen and Cooper, 1987; Dadamouny, 2009).

M. peregrina is the most drought tolerant species of the genus and native to rocky, semi-arid and arid areas of the Arabian Peninsula, Northeast Africa and Southwest Asia (Boulos 1999, Table S1), with a clear centre around the Red Sea with its gulfs. Further records come from Saudi Arabia, Palestine, Israel (El-Hadidi et al. 1992; Abdel-Wahab, 1995; Boulos 1999; Dadamouny 2009; Zaghloul et al. 2012), Sudan, Ethiopia, Somalia, Eritrea, and Djibouti to Yemen, Oman and United Arab of Emirates UAE (Padayachee and Baijnath, 2012; El- Awady et al., 2015). In Egypt, it is confined to the South Sinai Mountains and at the margin of its wadis (El- Hadidi et al. 1992, Dadamouny 2009) and the higher elevations of the north-facing slopes of the mountains around the Red Sea, from 1360 to 2187 m a.s.l (Zahran and Willis, 2009).

1.2. Putative threat factors for M. peregrina M. peregrina listed as a vulnerable in the national Red List of Egypt since 1992 (El-Hadidi et al. 1991, El- Hadidi et al. 1992). After 24 years, its populations are steadily declining (Dadamouny 2009), possible causes of threat are a rapid change towards a drier and more arid climate (Dadamouny and Schnittler 2016) and non- sustainable use (Abdel-Wahab 1995; Zaghloul et al., 2012). Both factors are likely to act synergistically, decreasing the overall amount of biomass palatable for livestock and thus increase pressure on the remaining populations (see Fig. S2). In the case of environmental stress, unmanaged human activities can drive a rare species to the brink of extinction (Moustafa, 2002; Zaghloul et al., 2012). We assume that with a recent trend towards a warmer and drier climate in the region, the tree is driven to the edge of its ecological niche. Decreasing annual precipitation hampers recruitment, and simultaneously flash floods caused by extreme rainfalls will destroy already established trees in wadi beds. In addition, overgrazing leads to severely reduced seeds set; most pods will be destroyed until the onset of the rainy season. All these factors will lead to a low establishment rate as recorded for the last three decades (Zaghloul et al., 2012). In Sinai, weather records between 1970 and 2014 showed a pronounced trend towards a warmer and drier climate (Dadamouny and Schnittler, 2016). Most likely, the trend in decreasing annual precipitation combined with increasing temperature (average increase +2.3 °C) will increase the level of aridity. In addition, extreme events, such as flash floods seem to become more frequent and unpredictable.

53 Publications Population performance and conservation of Moringa peregrina

1.3. Objectives of this study To conserve the remaining Moringa populations in Egypt, the first author initiated a monitoring program to study the population dynamics of M. peregrina (Dadamouny, 2009). His study focused on estimating the demographic status and revealing the relationship between tree size and age, in response to birth, aging and mortality rates for the remaining populations in South Sinai. To date, from the whole genus only M. arborea, which is native to (World Conservation Monitoring Centre, 1998), is included on the IUCN Red List (http://www.iucnredlist.org/). This study adds data for M. peregrina from Sinai to contribute more information relevant for a threat assessment of this species. Specifically, we aim to: (1) summarize the conservation status of the four largest populations in the area, using size distribution and annual increment in the different populations to forecast further performance; (2) reveal the possible interactions of threat factors, differentiating between direct human impact and indirect factors like climate change; and, (3) identify putative sites for facilitated recruitment. Finally, provide data to update the status of M. peregrina according to the recent version of IUCN Red List (IUCN ver. 12, 2016) based on a case study of the remaining populations at Sinai from 2006 to 2015. 2. Materials and Methods 2.1. Study area Extensive travel through all the major wadis of the Sinai Peninsula revealed that larger populations exist in only four wadis. The first (Meir, WM), is situated northeast of El-Tor, stretching over 18 km with a flat wadi bed up to 200–300 m wide; mainly covered by stones and rocky substrates (80%), with fine sand (20%) and dissected by channels and deltas due to the water erosion in rainy years (Dadamouny, 2009). The second wadi (Feiran, WF) represents one of the longest wadis in Southern Sinai (about 30 km long and 400–500 m wide). It is bounded by igneous and metamorphic mountains with different varieties of dikes (Said, 1990). Its main tributaries include many small wadis; W. El-Sheikh, Solaf, El-Akhdar, Nesrin, Tarr, Mekatab, and W. Alliat. The third wadi (Agala, WA) is a short and narrow tributary of W. Feiran ca. 4 km in length and 40–50 m wide (Dadamouny 2009). Its surface consists of the rocky substrate near the edges and gravel in the main water channel. The fourth wadi (Zaghra, WZ) is situated northeast of Saint Catherine, around 21 km from Dahab city, stretches over ca. 40 km long and is about 300–500 m wide (Dadamouny, 2009).

The wadis at Feiran and Agala harbor the largest Bedouins settlements of all studied wadis, accounting for ca. 10% of all inhabitants of South Sinai. These people depend on agriculture and pastoralism as their main source of income. W. Meir as well is home to ca. 700 Bedouins (Abou Jedr settlement). In W. Agala, Moringa trees occur in close vicinity to settlements and suffer therefore from grazing and cutting. W. Zaghra is near to the Coloured Canyon; here most Bedouins are centered on a settlement called Naj. Although the settlement is at some distance from the trees’ growth sites, plants are still exposed to drought and flash floods. 2.2. Census of Moringa populations Based on an initial survey started in 2000 for all Moringa trees at the four wadis (Dadamouny, 2009), a total census of all trees was repeated in 2006 and 2014. All performance parameters were measured in these two years; the tree ages were estimated in 2007 and 2015 after the approval of St. Catherine Protectorate to sample

54 Population performance and conservation of Moringa peregrina Publications

its wood. We measured or estimated the following parameters for tree performance: 1. tree height (m); 2. circumference and diameter at root collar (ground); 3. circumference at breast height cbh (cm) and tree diameter at breast height dbh; 4. crown cover area (m2); and, 5. tree vitality (estimated on a scale ranging from 1 to 5). In addition, two fitness parameters were recorded: 1. number of pods; and; 2. number of seeds per tree.

The tree diameter was calculated based on circumference (c) measurements (Franklin et al., 1988). Because Moringa has often multiple stems, we summed up cross area for all stems (A = πr2 with r = sqrt (c/π) and calculated the virtual diameter of a single stem from the total cross area. From the measurements carried out in 2006 and 2014, we calculated the increment area, excluding only a few (one to three) stems missed in one of the two surveys. From the mean of four diameters measured for the tree crown, we calculated the crown radius that is used to measure the crown cover area (πr2). For each site, a sketch map of all trees was drawn to facilitate relocation of the trees and to classify the surrounding landform. Finally, we estimated the tree vitality according to the following scale: 5 = vigorous growth excellent healthy plant; 4 = average tree, healthy but with yellow lower leaflets; 3 = still healthy but most lower leaves were yellow or shed; 2 = looking unhealthy or damaged, but twigs were mostly greenish, most leaves were yellow or shed; 1 = plant unhealthy, twigs turning were brown, nearly all leaves shed; 0= most twigs were brown or nearly dead, permanent wilting point reached (Dadamouny et al., 2012). From these measurements, we could derive four parameters describing tree performance; the differences 2014–2006 for height, dbh, cross area (increment), and vitality. To assess tree fitness, we counted the number of pods per tree (PPT), and from counting the number of seeds in five randomly collected pods we calculated the total number of seeds per tree (SPT). Similar to the tree performance, the differences in pod and seed set (2014–2006) were used as fitness parameters.

In summer of 2007 and 2015, we measured the circumference at ground (CAG). With the same method, we calculated the tree diameter (DAG). To relate diameter to tree age with a minimum impact on the trees, 117 dead branches of living trees were collected and annual rings were counted from these sections (Zoltai, 1975; Stockes and Smiley, 1986). A linear regression equation was employed to describe the relationship between age (number of annual rings) and dbh (diameter at breast height), forcing the regression curve to cut through the origin (Dadamouny, 2009).

Direct impact like grazing by livestock and cutting were estimated using scales. The first factor was assessed as the degree of browsing and the number of observed animals, mainly camels, modifying the scale (0 to 5) of Moustafa (2000) according to Abd El-Wahab (2003) and Dadamouny (2009) as follows: 5 = severe browsing, camels at the time of assessment present at the highest grazing rate; 4 = severe without the presence of camels but droppings abundant; 3 = intense signs of browsing on branches or pods, but no animals present and droppings regularly found; 2 = moderate signs of browsing, no animals present, droppings rare; 1 = low browsing; no animals were seen, droppings rare; 0 = no signs of browsing or animals. Similarly, cutting of fuelwood by humans was scaled from 0 to 5 with 5 = high degree of cutting and signs of wood burning; 4 = severe cutting without burning; 3 = moderate cutting with signs of burning; 2 = slight signs of cutting, no burning; 1 = low impact of only cutting; 0 = both cutting and burning absent. 55 Publications Population performance and conservation of Moringa peregrina

For each wadi, we collected the daily climate data from the nearest weather station (temperatures, total rainfall, relative humidity, wind speed, and air pressure at sea level from 1970 to 2014, see Dadamouny and Schnittler 2016). El-Tor, St. Catherine, and Dahab are the nearest stations to W. Meir, Feiran and Agala, and Zaghra, respectively. In addition, the following basic environmental parameters were recorded for each growth site: geographical position of the trees (hand-held GPS, UTM coordinates, precision 5–10 m); elevation in meters above sea level and above wadi bed, inclination, and exposition (measured clockwise in degrees from the North = 0°). The landform type (position in the wadi where a tree grows) was described as wadi-bed (WB), slopes (SL), foothills (FH), gorges (GR), and exposed rocks near to peaks (NP), modified from Moustafa and Klopatek (1995). 2.3. Evaluation of the remaining populations based on IUCN criteria We used all measured coordinates for the tree populations in Egypt and georeferenced coordinates using different distributions maps (see Table S1) in drawing the verified distribution map for M. peregrina excluding the sites that cultivated by M. oleifera. We calculated the Extent of occurrence (EOO) and Area of occupancy (AOO) using the program GeoCAT (http://geocat.kew.org/), considering all known records for each region (four wadis of Egypt, other African countries, and Arabian Peninsula with and without Oman and United Arab of Emirates UAE). Based on observations for the remaining M. peregrina in Egypt and the estimated areas (EOO and AOO), we applied the IUCN criteria for local and worldwide evaluation in order to update the current status after the last local evaluation by El-Hadidi et al. (1992). 2.4. Data treatment Data were first evaluated by descriptive statistics, employing the Anderson-Darling normality test, Pearson correlation analysis and simple linear regression (Minitab, 2014; SPSS, 2015). Correlations between environmental and tree parameters were calculated with Pearson’s linear correlation. The significance between the first assessment in 2006 and the second in 2014 of each environmental parameter and tree performance / fitness was tested using Wilcoxon and T-test. Also, the significance between the four wadis in each assessment was tested using one-way analysis of variance (ANOVA) to investigate if these parameters differed significantly between wadis (Henson 2015). Relationships between environmental parameters and tree performance / fitness were explored by multivariate analysis, employing nonmetric multidimensional scaling (NMDS). Principal coordinate analysis (PCO) was used to examine the response of tree parameters to environmental factors (McCune, 2002; Kent, 2012). The analyses were conducted with R (R Core Team, 2012). 3. Results 3.1. Changes in population size The locations of the four investigated populations at the Sinai Peninsula are shown in Fig. 1. The number of labelled M. peregrina trees in the total population declined from 448 trees in 2000 to 423 trees in 2006, then 393 trees in 2014 finally to 390 in 2015. W. Meir has the highest number of Moringa trees in Sinai (2006: 249; 2014: 242 trees). In W. Feiran, Moringa inhabits rock crevices on high elevations exceeding 800 m.a.s.l, and these sheltered growth sites seemed to prevent major losses (2006: 47, 2014: 45 trees). Losses in W. Agala were higher in comparison to its population size (2006: 40, 2014: 35 trees), and the situation was similar in W. Zaghra (2006: 87, 2014: 71 trees). After the rainfall of 2010 and 2013, the vitality scale improved in 2014 (3.8±81 at .... 56 Population performance and conservation of Moringa peregrina Publications

Fig. 1. (A): Location of the region in Egypt with the studied wadis at Sinai Peninsula (B), both images Google Earth, 22 September 2015. (C–F): Satellite images (Google Earth, 25 February 2009) with the superimposed location of the studied trees, circles are scaled according to dbh. UTM coordinates for the lower left edges of the images (36R) are C. W. Meir (3148000N, 572500E), D. W. Agala (3174000N, 560400E), E. W. Feiran (3175700N, 562800E), and F. W. Zaghra (3170400N, 626200E). Red circles represent the trees recorded in 2006 but disappeared in 2015. Saplings recorded in 2014 are marked by a green symbol (N), but all disappeared again in 2015 (For interpretation of the signatures in colour in this figure legend, the reader is referred to the web version of this article).

WM, 3.4±0.65 at WA, 3.1±0.69 at WF, and 3.9±0.75, Table 1). Moringa trees at all four wadis showed a low to missing recruitment during the last three decades (the youngest tree was estimated to be 27 years). Figure 2 shows the size structure of the four populations (figures from 2006 and 2014, 393 investigated trees from

57 Publications Population performance and conservation of Moringa peregrina four wadis). Except for the 12 saplings observed as new in 2014 but dead in 2015, no trees below 5 cm dbh were recorded, and only four trees belonged to the second size class (5–<10 cm dbh; three at W. Feiran and one at W. Zaghra).

Table 1. Summary statistics of the investigated abiotic parameters (elevation above sea level, inclination or slope, exposition aspect, browsing, and cutting intensity, mean annual temperatures (2006–2014), and mean of annual precipitation in the period 2006 to 2014) and the investigated biotic (tree) parameters for M. peregrina populations (mean ± SD); in 2006 and for the same trees in 2014, and the mean differences (2014–2006); means of annual tree radius increment (mm) and area at breast height (cm2), for trees in wadi Meir (WM), w. Agala (WA), w. Feiran (WF), and w. Zaghra (WZ). Missed trees in 2014 were excluded from the calculations of 2006.

Environmental parameters WM WA WF WZ Elevation a.s.l [m] 460 ± 37.2 692 ± 42.3 702 ± 40.6 641 ± 94.1 Inclination [slope, o] 23.8 ± 11.0 45.3 ± 4.7 55.4 ± 8.3 33.2 ± 8.5 Exposition [o] 160.12 ± 6.7 204.94 ± 20.7 155.60 ± 15.1 201.63 ± 15.11 Browsing [scale 1–5] 3.0 ± 1.6 3.9 ± 0.9 0.9 ± 0.3 0.5 ± 0.8 Cutting [scale 1–5] 2.1 ± 1.3 3.2 ± 1.2 3.2 ± 0.9 2.6 ± 1.6 Mean Temp. [◦C] 25.5 ± 2.1 18.5 ± 1.6 18.5 ± 1.6 24.8 ± 1.5 Precipitation [mm] 8.3 ± 7.8 13.4 ± 13.4 13.4 ± 13.4 40.5 ± 37.7 Trees Parameters 2006 No. of trees 242 35 45 71 Height [m] 7.0 ± 3.7 4.7 ± 2.5 4.8 ± 2.6 8.1 ± 3.4 Dbh [cm] 38.4 ± 26.6 19.6 ± 12.0 18.2 ± 12.7 39.2 ± 21.3 Crown [m2] 20.23 ± 18.81 9.63 ± 6.34 11.60 ± 11.1 26.30 ± 20.67 Vitality [1–5] 2.9 ± 0.7 3.0 ± 0.6 2.7 ± 0.6 3.0 ± 0.3 Pods [No.] 8.9 ± 13.9 5.2 ± 11.4 5.1 ± 7.1 31.8 ± 35.9 Seeds [No.] 91.8 ± 148.5 54.9 ± 120.9 53.8 ± 72.7 348.6 ± 386.7 Trees Parameters 2014 No. of trees 242 35 45 71 Height [m] 8.0 ± 3.7 5.6 ± 2.6 6.0 ± 2.8 9.6 ± 3.4 Dbh [cm] 42.0 ± 26.9 23.7 ± 11.9 22.1 ± 13 43.5 ± 21.4 Crown [m2] 21.5 ± 18.9 11.3 ± 6.45 12.9 ± 11.1 28.5 ± 21.0 Vitality [1–5] 3.6 ± 0.81 3.4 ± 0.65 3.1 ± 0.69 3.9 ± 0.75 Pods [No.] 18.7 ± 24.6 5.46 ± 10.4 12.0 ± 16.9 43.6 ± 54.5 Seeds [No.] 174.0 ± 232.2 48.5 ± 93.9 109.0 ± 155.3 472.7 ± 584.7 Annual radius increment [mm] 1.9 ± 0.4 2.3 ± 0.5 2.2 ± 0.5 2.3 ± 0.4 Annual increment area at bh [cm2] 29.6 ± 23.9 17.2 ± 8.52 16.3 ± 12.4 34.9 ± 19.5

3.2. Importance of environmental parameters for population decline Summary statistics of the four Moringa populations (mean±SD) for all investigated abiotic (environmental) and biotic (tree performance and fitness) parameters are given in Table 1. Figure 1, tracing the exact locations of trees, reveals them to be found mostly at the margin of the wadis, where the trees are still supplied by water but somewhat sheltered from flash floods. The maximum / mean elevation above the wadi surface were 198 / 34 m for W. Meir, 297 / 97 m for W. Agala, 249 / 74 m for W. Feiran, and 152 / 18 m for W. Zaghra. Tree vitality showed a weak negative correlation with elevation above wadi surface (-0.36) and with the slope (- 0.22). The highest inclination of the slopes was recorded for the sites in W. Feiran (60°), sparing effectively these trees from grazing (mean intensity 0.9±0.3). A similar situation was found in the trees of W. Zaghra (Table 1), whereas trees in W. Feiran and W. Agala suffer from cutting for fuelwood (scale figures 3.2±1.2 and 3.2±0.9, respectively). Other threat factors are drought and flash floods. As shown in Table 1, W. Zaghra received the highest rainfall in the period 2006 to 2014 but was hit by a flash flood in 2010, which eradicated 11 trees at two sites (Fig. 1, red dots represent vanished trees). However, after the strong rainfalls in 2010 twelve saplings (all < 1m height) were recorded in 2014 (see green symbols in Fig. 1) in three wadis, but all were found dead in 2015.

58 Population performance and conservation of Moringa peregrina Publications

Fig. 2. Histogram of dbh values in classes of 5 cm dbh for all studied M. peregrina trees in 2014 showed the absence of trees in the lowermost dbh class (0– < 5 cm) at the four wadis in Sinai (A: W. Meir, B: W. Agala, C: W. Feiran, and D: W. Zaghra).

3.3. Recruitment and age estimates Based on the monitoring of the remaining populations from 2006 up to 2015, all four populations show missing recruitment and the dominance of old trees (Fig. 2 & Fig. 3). Seed germination exceeds 90%, but the chance for seeds to survive in the soil bed until a rain event is low. A soil seed bank seems not to exist (only one in ten trees could keep some pods on higher branches for the next year). For dbh values up to 16 cm, we obtained a fairly good linear correlation between dbh and tree age, the latter estimated as the number of annual rings (age = 2.8 × dbh, R2 = 0.562). But, bark thickness should be subtracted from measured dbh. Based on our measurements of bark thickness from 117 cross sections of stems, we found a relationship: bark thickness = 0.116 × Radius (R2 = 0.253), resulting in a corrected relationship age = 5.68 × (radius – bark thickness), R2 = 0.56). The oldest Moringa trees were estimated for W. Zaghra (see dbh classes in Fig. 2 and estimated mean tree age in Fig. 3). All performance and fitness parameters were negatively correlated with tree age. 3.4. Change in tree performance Figure 3 shows also the means for the absolute values in 2014 for six performance (height, dbh, cross area, age, Crown diameter, and vitality) and two fitness parameters (numbers of pods and seeds); in addition, the mean differences 2014–2006 are shown. Trees at Wadi Zaghra performed best (highest differences in all parameters), followed by W. Meir. This pattern essentially repeats for the two fitness parameters; pod and seed set, but Wadi Agala sticks out, producing after eight years no more seeds than 2006. Figure 4 reveals that W. Feiran which has on average the youngest trees, shows as well the highest mean increment in stem diameter …… 59 Publications Population performance and conservation of Moringa peregrina

Fig. 3. Values of mean ± SD of measured tree parameters in 2014 (white bars and left Y-axis) and the mean differences between 2014 and 2006 (mean ± SD, gray bars and right Y-axis) in the four wadis (Meir: WM, Agala: WA, Feiran: WF, and Zaghra: WZ). Shown are six performance parameters: A. Tree height [m], B. dbh [cm], C. annual increment area [cm2], D. estimated age [years] E. crown area [m2], F. vitality [scale 1–5] and two fitness parameters: G. number of pods per tree, H. number of seeds per tree. Significant differences between 2006 and 2014 (for tree age between 2007 and 2015) according to a T-test are indicated as * (p < 0.001) and ** (p > 0.05).

60 Population performance and conservation of Moringa peregrina Publications

Fig. 4. Box plot showing mean increment for the four populations of Moringa peregrina, measured as the dbh ratio between 2014 and 2006. The highest increment was recorded for the trees of W. Feiran (WF), followed by trees of W. Meir (WM), W. Zaghra (WZ) and W. Agala (WA).

Table 2. Seed set of Moringa peregrina in 2006 and 2014, shows the numbers of trees with seed set in single years and in both years, and the proportions on the respective populations (as counted in 2014), and differences in seed set.

Trees with seeds Differences in seed set 2006 % 2014 % 2006 or 2014 % Positive % Negative % Total 210 53 216 55 279 71 105 27 174 44 W. Meir 123 51 141 58 178 74 64 26 114 47 W. Agala 12 34 11 31 18 51 9 26 9 26 W. Feiran 21 47 17 38 23 51 7 16 16 36 W. Zaghra 54 76 47 66 60 85 25 35 35 49

ratio dbh 2014 / dbh 2006; followed again by W. Meir. Based on estimated tree ages, the youngest tree was estimated to be 27 years at W. Feiran, 32 at W. Agala, 36 at W. Meir, and 30 years at W. Zaghra. 3.5. Seed set We found high fluctuations in numbers of produced pods (seeds) between our two censuses. Looking at the total population, 63 Moringa trees produced seeds in 2006 but not in 2014, while 69 trees produced seeds in 2014 but not in 2006. More than one-fourth (114 out of 393) trees did not produce seeds in either year (Table 2), with most located at the wadis Agala and Feiran. We recorded the highest number of pods (340) for a tree in W. Zaghra which was estimated to produce about 3200 seeds. But for this wadi, we found as well the highest differences during 2006–2014 in the mean pod (11.8 ± 40.6) and seed (124 ± 429) numbers. A single tree produced 11% of the total population pods in 2006 and 20% in 2014. Figure 3 shows differences in seed set between 2006 and 2014; negative values indicate lower seed set in 2014. Around 27% of all trees in the pooled population produced fewer seeds in 2014 (Table 2). Although the trees of W. Zaghra showed the best overall seed set, 35% of all seed-producing individuals (25 trees) produced fewer seeds in 2014, compared with 16% of the seed-producing individuals in W. Feiran (7 trees). Half of M. peregrina trees are located on south-facing slopes (204 trees or 52%), including nearly all trees showing seed set (Table S5). At the same wadis, trees growing on north-facing slopes developed a high number of flowers and pods, but seed number was higher for trees growing on south-facing ones (Tables S5 and S6). Only about 40% of all trees set seeds in a given year.

61 Publications Population performance and conservation of Moringa peregrina

3.6. Interactions between threat factors Naturally, performance parameters were highly auto-correlated with each other (for instance, the correlation coefficient was 0.602 for height and dbh and 0.574 for height and increment area). However, the correlation between the increment and number of seeds was much lower (0.189). Differences in all tree performance and fitness parameters between 2006 and 2014 were statistically significant (P<0.000, Wilcoxon and paired T-test). Moreover, all performance and fitness parameters differed significantly between the four wadis (P=0.000 for a single-factor ANOVA).

NMDS and PCO were first conducted separately for the years 2006 and 2014. Since the 2006 data resulted in similar ordination patterns and NMDS principally showed congruency to PCO, only PCO ordinations for 2014 data are shown in Figure 5. Species scores (colored according to wadi type) reveal wadis Meir and Zaghra (the much warmer wadis) to have better performing trees, and precipitation points in this direction. W. Zaghra is the wadi receiving the highest annual precipitation.

Fig. 5. Principal coordinate analysis of tree performance, including four descriptors of performance (tree dbh, height, vitality, and annual increment) against nine environmental parameters (elevation a.s.l, inclination, exposition, landform, mean annual precipitation, mean annual temperature, browsing, and cutting, and GPS coordinates UTM; Easting, and Northing) for the year 2014. Significant parameters (abbreviated parts of names written in Italics, P = 0.05) are displayed as arrows pointing towards the direction of effect. Sample scores are represented by symbols colored according to the location of trees in different wadis; figures at the upper left denote the eigenvalue of the first two axes.

3.7. Threat assessment By applying the IUCN criteria using the collected data for the number of individuals, coordinates, and annual field observations resulted in an endangered status at least first in the local assessment for M. peregrina trees in Egypt (see Table S3 for worldwide and local evaluation). The updated status for the trees in Egypt is also explained in Table S4. For the four recorded populations, the extent of occurrence was 1,020.3 km2 and area of occupancy was 36 km2 (including the area between the wadis, but for the total wadis was only 18.56 km2). For each wadi, the values of EOO/AOO were 2.201/16 km2 in WM, 0.155/0.540 km2 in WA, 0.219/0.270 in WF, and 3.750/1.750 km2 (with an outlier tree) and only 0.444/1.500 km2 (without outliers) in WZ (see Fig. 62 Population performance and conservation of Moringa peregrina Publications

S3). The assessment revealed that the remaining M. peregrina populations belong in an IUCN Red List Threatened Category Endangered (EN) according to the criteria (A3, B1, B2, C, and C2) and Vulnerable (VU) according to the criteria (A1, C1, D, D2). 4. Discussion 4.1. Causes of threat In Sinai, an increase in mean annual temperature by +2.3°C from 1970 to 2014 (Dadamouny and Schnittler, 2016) seems to be underlaid with increasing fluctuations in precipitation. Severe droughts lasting for several years can be interrupted by extreme rainfalls, delivering most of the annual precipitation during one to three days only. Due to the extremely arid climate, M. peregrina is confined to wadis and the trees are well accessible and thus under high pressure of human activities (e.g. cutting or uprooting for fuel, collections for medicinal uses, grazing by livestock, a collection of the nutritious seeds). Direct impact and (indirectly) climate change seemed to cause recruitment failure within the last three decades. In addition, seed set fluctuates between years, making a coincidence of reasonable seed set and one, even better two, rainy years a very rare event. According to Abd El-Wahab 1995, Batanouny et al. 2009, Zahran and Wills 2009, and others, the wild Moringa populations have been reduced to a few populations distributed in South Sinai, the Eastern Desert and Gebel Elba in south-east Egypt and are on a continuous decline. The west-facing escarpments of Gebel Serbal in Sinai were once rich in M. peregrina trees, which grow on rocky slopes near springs (Danin, 1999) but nearly all specimens died during a series of extremely dry years. Bedouins on Sinai depend on animal husbandry as a resource of income (Heneidy and Halmy, 2009). But, our multivariate analyses revealed a comparatively low effect of direct human impact, such as cutting and grazing pressure. Moringa trees have a high palatability and lack spines, in contrast to most other wild trees (e.g. Acacia sp.). The highest grazing pressure is exhibited during the summer season when annual plants have disappeared. At this time (June to August) M. peregrina produces pods, which are a valued fodder for the Bedouins' animals, and these losses in seeds set are likely to be one, but not the only, reason for the virtually lacking recruitment of the populations. The recruitment problem is worsened by the observation (Young, 1984) that solar radiation increases floral development in M. peregrina. All reproductive trees (with the highest number of flowers) were recorded on the south-facing slopes. Naturally, a south-facing slope at this exposition (135–225°) experiences a higher radiation, has warmer air and soil (Barbour et al., 1987) but is more susceptible to drought. Abd El-Wahab (1995) and Dadamouny (2009) recorded that most of M. peregrina trees grow in between rock crevices, on cliffs, and at the base of hills with very rugged topography. The elevation of these hills ranges between 700 to 800 m a.s.l. The trees at a high elevation showed a low vitality, where the tree vitality and its position above the wadi bottom are negatively correlated (-0.36). In addition, trees at elevated positions were usually smaller, and can be more easily uprooted by storms, and these trees suffered the highest losses from 2006 to 2014. We found that M. peregrina trees located at the wadi bed expressed a better vitality and the highest growth rates. But, at this position the trees face a higher risk of over-grazing and being eradicated by rare but severe flash floods. Undoubtedly, the critical phase in the life of M. peregrina is the survival of newly established seedlings (seeds usually germinate after winter rainfalls). Conservation efforts should be

63 Publications Population performance and conservation of Moringa peregrina

directed to these saplings, guiding them especially during the summer months to give these saplings a chance to be established in its natural habitats. We lack long-term data to decide if the highly fluctuating seed set in M. peregrina is a natural phenomenon or already a consequence of the drier and warmer climate in the region. However, this trend could lead to a vicious circle of lacking seed set, low sapling survival, and eradication of the old trees by occasional flash floods. 4.2. Conservation status Effective conservation of rare plant species requires a detailed understanding of their distributions and habitat requirements to identify conservation targets (Schnittler and Günther, 1999; Crain et al., 2011; Crain et al., 2014). Gorbunov et al. (2008) stated the preservation of individual species and plant communities (in-situ) is preferable than simulating habitats (ex-situ). However, in many regions, human pressure is simply too high to allow the preservation of plants under natural conditions. This applies as well to the remaining populations of M. peregrina in South Sinai, which face an acute risk of disappearing completely in the near future.

Further studies should prove if a classical matricial approach for determining lambda like that of Pisanu et al. (2012), is feasible to be used in forecasting the fate of our populations. But, we currently have only data from two years: 2006 and 2014 (for all parameters) and 2007 and 2015 (for tree age). Given the perhaps naturally fluctuating seed set and recruitment, this rules out a transition matrix approach, since the data will not allow a reliable forecast for a long-lived organism like this tree. At present, we can only state the absence of recruitment within the last three decades.

Based mainly on the EOO, our threat assessment nevertheless suggests the status Endangered, which is more severe than the last assessment (vulnerable) by El-Hadidi et al. (1992). In addition, we assume both EOO and AOO to be overestimated because the tree does not occur in pure sand deserts, which make up for most of the inner parts of the Arabian Peninsula. Likewise, the tree avoids high mountainous areas of Sinai (like St. Catherine mountains, where the temperature is too low) and, due to frequent flash floods, trees are virtually absent from large wadis (like W. Watir). Therefore, we recommend including M. peregrina in the IUCN Red List of Threatened Species.

4.3. What are suitable sites for future recruitment? Following Coomes and Allen (2007) and Wyckoff and Clark (2005), exposed crown area (average R2 = 0.69 for nine species) and tree size alone (average R2 = 0.68 for 21 species) are an excellent predictor of future tree growth. Among the four wadis investigated, trees of W. Zaghra exhibited the highest increase in tree size, but tree performance was better for the trees of W. Feiran (trees of this wadi are on average younger than those of W. Zaghra). In general, trees growing in the wadi-bed were most vital. Many studies (e.g. Sorman and Abdulrazzak, 1993; Springuel et al., 2006; Wiegand et al., 1999; Wiegand et al., 2000) proved that wadi beds are the most suitable sites, and sometimes the only chance for a tree to survive. As shown by the multivariate analyses, trees seem to tolerate overuse by cutting and grazing to a certain extent, whereas the risk of its eradication by flash floods is hard to estimate but certainly higher. Considering this apparent trade-off (best grow conditions at the wadi beds, but highest probability of impacts by grazing, cutting, and flash floods), the 64 Population performance and conservation of Moringa peregrina Publications

best sites for future recruitment may be the lower wadi slopes, just out of the reach of possible flash floods. The density of human population is an additional consideration: the trees of W. Feiran and W. Agala show the highest performance but have the worst conservation status due to the high density of the Bedouins population in these two wadis. The less affected is W. Meir which has many trees, assuming an intermediate growth position at the lower wadi slopes. Therefore, W. Meir may offer the most suitable sites for future facilitated recruitment. 4.4. Proposed solutions Looking at the currently high human pressure, ex-situ conservation should be one of the first measures. Prendergast (1994) preserved M. peregrina seeds collected during four expeditions to Oman, and seeds are conserved as well in the National Center for Agricultural Research and Extension in Jordan. Looking for in-situ measures, we favor facilitated propagation of M. peregrina by seeds and stem cuttings in its natural habitats. Beside this, conservation efforts should focus on the regeneration of destructed vegetation to decrease grazing pressures on the tree. A conservation program should explore several approaches, e.g. by through collecting seeds from both old (in W. Zaghra) and young trees (in W. Feiran), or subsequent propagation either by seeds or stem cuttings in natural habitats (preferably on the mountain foothills). For reintroduction experiments, W. El-Gemal-Hamata National Park (see EEAA, 2014) may be one of the most suitable sites, along with the Red Sea Mountains. As a second approach, the temporary fencing of especially young M. peregrina trees in South Sinai during early establishment and growth may help to enhance recruitment. So far, we do not have data on the genetic structure and adaptation of the naturally fragmented M. peregrina populations. Therefore, studies on its reproductive biology and the effect of population size and genetic relatedness of the remaining Moringa trees on reproductive success are in needed. 5. Conclusion Moringa peregrina is a traditionally highly valued, endangered tree species and may thus be well suited to represent a flagship species for a conservation program. The evaluation of tree performance parameters (tree height, tree circumference/diameter, crown cover area, and vitality scale) indicates that climate change and human impacts affect synergistically growth and seed set of Moringa trees, and contributed to the population decline by 15% within nine years (2006–2015). However, in a region with naturally fluctuating rainfalls, it is difficult to ascertain trends in Moringa tree recruitment. Habitat choice for the tree reflects a trade-off: wadi beds are the best growth sites regarding water supply, but are most affected by direct human impact and flash floods. Similarly, seed set, but as well water stress, is better on the south than on north-facing slopes. Wadi Meir seems to offer best conditions for future recruitment on its lower slopes. Conservation programs; like short-term fencing during establishment periods, sustainable use, and propagation of the tree, are necessary, to avoid its local extinction. Also, ex-situ measures (collecting seeds and/or reintroduce saplings) seem to be promising. As the causes of threat identified by us continue to act, further monitoring is needed to assess, the effect of declining population size and the dominance of old individuals on survival and genetic diversity of the remaining populations at the Sinai Peninsula. Acknowledgments First author thanks the German Academic Exchange Service (DAAD Bonn) for funding (57076387-GERLS- 91541411). Appreciation extends to DAAD Cairo and Ministry of Higher Education Egypt. We would like to

65 Publications Population performance and conservation of Moringa peregrina thank all staff of Environment Research Centre, at St. Catherine, Egypt for their support during data collection. The author's appreciation extends to St. Catherine Protectorate authorities for facilitating the field work. Thanks to Mohamed Abou Wadid and Ouda Ahmed (the trees guards, St. Catherine Protectorate) for their efforts in protecting Moringa trees in W. Zaghra and W. Meir. Thanks to Mrs. Riham Elghandour and Mr. Robert Weigel, Institute of Botany and Landscape Ecology, Greifswald for helping in computer work.

References Abd El-Wahab, R.H. (1995). Reproduction Ecology of Wild Trees and Shrubs in Southern Sinai, Egypt. M.Sc. Thesis, Botany Department, Faculty of Science, Suez Canal University, Ismailia, Egypt. p. 110. Abd El-Wahab, R.H. (2003). Ecological Evaluation of Soil Quality in South Sinai, Egypt, Ph.D. Thesis, Botany Department, Faculty of Science, Suez Canal University, Ismailia, Egypt. p. 206. Al-Dabbas, M. M., Ahmad, R., Ajo, R.Y., Abulaila, K. H., Akash, M., & Al-Ismail, K. (2010). Chemical composition and oil components in seeds of Moringa peregrina (Forssk) Fiori. Crop. Res., 40(1–3), 161–167. Al-Kahtani, H.A., & Abou-Arab, A.A. (1993). Comparison of physical, chemical, and functional properties of Moringa peregrina (Al-Yassar or Al-Ban) and soybean proteins. Cereal Chemistry, 70(6), 619–626. Asghari, G., Palizban, A., Bakhshaei, B. (2015). Quantitative analysis of the nutritional components in leaves and seeds of the Persian Moringa peregrina (Forssk.) Fiori. Pharmacognosy Research, 7(3), 242–248 doi:10.4103/0974- 8490.157968 Barbour, M. G., Burk, J.H., Pitts, W. D. (1987). Terrestrial Plant Ecology. California: Benjamin Cummings, Menlo Park, CA. p. 634. Boulos, L. (1999). 'Flora of Egypt.' Vol. I (Azollaceae – Oxalidaceae). Al hadara Publishing, Cairo, Egypt, p. 419. Bruna, E. M., Ribeiro, M. B. (2005). The Compensatory responses of an understory herb to experimental damage habitat- dependent. American J. Botany, 92(12), 2101–2106. Coomes, D. A., & Allen, RB. (2007). Effects of size, competition, and altitude on tree growth. Ecology 95, 1084–1097 doi:10.1111/j.1365-2745.2007.01280.x Crain, B. J., White, J. W., & Steinberg, S. J. (2011). Geographic discrepancies between global and local rarity richness patterns and the implications for conservation. Biodivers. Conserv., 20, 3489–3500 doi:10.1007/s10531-011-0137-6. Crain, B. J., Cuervoa, A. M. S., White, J. W., & Steinberg, S. J. (2014). Conservation ecology of rare plants within complex local habitat networks. Fauna & Flora International, Oryx 1–8 doi:10.1017/S0030605313001245. Dadamouny, M. A. (2009). Population Ecology of Moringa peregrina growing in Southern Sinai, Egypt. M.Sc. thesis, Environmental Sciences, Bot. Dept., Faculty of Science, Suez Canal University, Ismailia, Egypt. p 205. Dadamouny, M. A., Zaghloul, M. S., & Moustafa, A. A. (2012). Impact of Improved Soil Properties on Establishment of Moringa peregrina Seedlings and a Trial to Decrease its Mortality Rate. Egyptian J. Botany, 52(1), 83–98. Dadamouny, M. A., & Schnittler, M. (2016). Trends of climate with rapid change in Sinai, Egypt. J. of water and climate change, 7(2), 393–414. IWA doi: 10.2166/wcc.2015.215. Dehshahri, S., Wink, M., Afsharypuor, M., Asghari, G., & Mohagheghzadeh, A. (2012). Antioxidant activity of methanolic leaf extract of Moringa peregrina (Forssk.) Fiori. Research in Pharmaceutical Sciences, 7(2), 111–118. Danin, A. (1999). Desert rocks as plant refugia in the Near East. Avinoam Danin. The Botanical Review, 65, 12–93. EEAA Egyptian Environmental Affairs Agency (2014). Official Record for Wadi El-Gemal-Hamata, Ministry of Environment, Egypt. http://www.protectedplanet.net/wadi-el-gemal-hamata-national-park accessed December 2015. El-Alfy, T. S., Ezzat S. M., Hegazy, A. K., Amer, A. M., & Kamel, G. M. (2011). Isolation of biologically active constituents from Moringa peregrina (Forssk.) Fiori. (Family: Moringaceae) growing in Egypt. Pharmacogn Mag, 7(26), 109–115. El-Hadidi, M. N. (1992). A historical flora of Egypt, a preliminary survey. In Friedman R, Adams B (eds.). The followers of Horus, studies dedicated to MA Hoffman 1944–1990. Oxford 144–155. El-Hadidi, M. N., Batanouny, K.H., & Fahmy, A.G. (1991). The Egyptian Plant Red Data Book, Volume I, Faculty of Science, Cairo University, United Nations Environment Programme (UNEP) Geneve and Nairobi, p. 226. El-Hadidi, M. N., Abd El Ghani, M.M., & Fahmy, A.G. (1992). The Plant Red Data Book of Egypt. Woody Perennials. Palm Press and Cairo University Herbarium. ISBN - 977-5089-04-2. Franklin, J. F., & DeBell, D.S. (1988). Thirty-six years of tree population change in old-growth Pseudostsuga- Tsuga forest. Canadian J Forest Research, 18, 633–639.

66 Population performance and conservation of Moringa peregrina Publications

Gorbunov, Y. N., Dzybov, D. S., Kuzmin, Z. E., Smirnov, I. A. (2008). Methodological recommendations for botanic gardens on the reintroduction of rare and threatened plants. Grif & Co, Tula Hartwell, J. L. (1971). Plants used against cancer: a survey. Lloydia's 34(1), 103–1160. Heneidy, S. Z., & Halmy, M. (2009). The nutritive value and role of Panicum turgidum Forssk. in the arid ecosystems of the Egyptian desert. Acta Botany Croatica, 68(1), 127–146. Henson, R. N. (2015). Analysis of Variance (ANOVA). In Arthur W. Toga, editor. Brain Mapping: An Encyclopedic Reference. Elsevier Academic PressAcademic Press 1, 477–481. IUCN (1992). National Red List: Moringa peregrina, http://www.nationalredlist.org/species- information/?speciesID=129124, accessed 24 December 2015. Kent, M. (2012). Vegetation Description and Data Analysis. Wiley-Blackwell, p. 414. Koheil, M.A., Hussein, M.A., Othman, S.M., & El-Haddad, A. (2011). Anti-inflammatory and antioxidant activities of Moringa peregrina seeds. Free Radicals and Antioxidants, 1(2):49–61. Lalas, S., Gortzi, O., Athanasiadis, V., Tsaknis, J., & Chinou I. (2012). Determination of antimicrobial activity and resistance to oxidation of Moringa peregrina. Seed oil Molecules, 17, 2330–2334. McCune, B., & Grace, J.B. (2002). Analysis of Ecological Communities. MjM Software Design: Gleneden Beach, Oregon. p. 300. MINITAB (2014). Meet MINITAB 17 for Windows, a concise guide to getting started with Minitab software [online http://www.minitab.com/en-us/products/minitab/] accessed Oct 2015. Moustafa, A.A. (2000). Impact of grazing on the vegetation of South Sinai, Egypt. In Sustainable Land Use in Desert, Springer, Germany 218–228. Moustafa, A. A. (2002). Demography and Age dating of Acacia in Saint Katherine Protectorate. Final Report, (SEM 04/220/016A - Egypt), Saint Catherine Protectorate Development Project, EEAA. Moustafa, A. A., & Klopatek, J. M. (1995). Vegetation and landforms of the St. Katherine area, Southern Sinai, Egypt. Arid Environments, 30, 385–395. Nawash, O. S., & Horani, A. A. (2011). The most important medicinal plants in Wadi Araba desert in South West Jordan: a review article. Advances in Environmental Biology, 418–426. Olson, M. (2002). Combining Data of DNA Sequences and Morphology for a Phylogeny of Moringaceae (Brassicales). American Society of Plant Taxonomists. Journal of Systematic Botany, 27(1), 55–73 Olson, M.E., Carlquist, S. (2001). Stem and root anatomical correlations with life form diversity, ecology, and systematics in Moringa (Moringaceae). Botanical Journal of the Linnean Society, 135(4), 315–348. Owen, S. N., & Cooper, S. M. (1987). Palatability of woody plants to browsing in a South African savanna. Ecology, 68, 319–331. Pisanu, S., Farris, E., Filigheddu, R., & García, M. B. (2012). Demographic effects of large, introduced herbivores on a long-lived endemic plant. Plant Ecology, 213(10), 1543–1553. Prendergast, H. D. V. (1994). Tree seeds from the Sultanate of Oman in the Kew Seed Bank. FAO/IBPGR Plant Genetic Resources Newsletter, 97, 37–42. R Core Team (2012). R: A language and environment for statistical computing. R Foundation for Statistical Computing, Vienna, Austria. ISBN 3-900051-07-0, URL http://www.R-project.org/ Said, R. (1990). The geology of Egypt, Elsevier Pub Amsterdam, p. 734. Schnittler, M., & Günther, K. F. (1999). Central European vascular plants requiring priority conservation measures–an analysis from national Red Lists and distribution maps. Biodiversity and Conservation, 8(7), 891–925. Soltan, M. M., & Zaki, A. K. (2009). Antiviral screening of forty-two Egyptian medicinal plants. Ethnopharmacology, 126(1), 102–107. Somali, M. A., Bajneid, M. A., & Al-Fhaimani, S. S. (1984). Chemical composition and characteristics of Moringa peregrina seeds and seeds oil, J. the American Oil Chemists' Society, 61(1), 85–86. Sorman, A. U., & Abdulrazzak, M. J. (1993). Infiltration - recharge through wadi beds in arid regions. Hydrological sciences journal, 38(3), 173–186. Springuel, I., Sheded, M., Darius, F., & Bornkamm, R. (2006). Vegetation dynamics in an extreme desert wadi under the influence of episodic rainfall. Polish Botanical Studies, 22, 459–472. SPSS, IBM Corp, & Armonk, N. Y. (2015). IBM SPSS Statistics for Windows, Version 22.0. IBM Corp. Released Stockes, M. A., & Smiley, T. L. (1986). An introduction to tree-ring dating. Chicago: University of Chicago Press. Wiegand, K., Jeltsch, F., & Ward, D. (1999). Analysis of the population dynamics of Acacia trees in the Negev desert, Israel with a spatially-explicit computer simulation model. Ecological Modelling, 117, 203–224. Wiegand, K., Jeltsch, F., & Ward, D. (2000). Do spatial effects play a role in the spatial distribution of desert-dwelling Acacia raddiana? Vegetation Science, 11, 473–484.

67 Publications Population performance and conservation of Moringa peregrina

World Conservation Monitoring Centre (1998). Moringa arborea. The IUCN Red List of Threatened Species 1998: e.T32641A9720645. http://dx.doi.org/10.2305/IUCN.UK.1998.RLTS.T32641A9720645.en accessed 22 October 2015. Wyckoff, P. H., & Clark, J. S. (2005). Tree growth prediction using size and exposed crown area. Can. J. For. Res., 35, 13–20 doi:10.1139/X04-142. Young, T. P. (1984). Solar Irradiation Increases Floral Development Rates in Afro-Alpine Lobelia telekii. Biotropica 16(3), 243–245. Zahran, M. A., & Willis, A. J. (2009). The vegetation of Egypt, 2nd Ed., London: Chapman and Hall, p. 437. Zaghloul, M. S., Moustafa, A. A., & Dadamouny, M. A. (2012). Age structure and population dynamics of Moringa peregrina, an economically valuable medicinal plant. Egyptian J. Botany, 52(2), 499–519. Zoltai, S.C. (1975). Tree-ring record of soil movements on permafrost. Arctic and Alpine Research, 7, 331–340.

Received 24 April 2016 Received in revised form 24 August 2016 Accepted 25 August 2016

68

Appendix III

Population performance of Moringa peregrina (Forssk.) Fiori (Moringaceae) at Sinai Peninsula, Egypt in the last decades: Consequences for its conservation

∗ Mohamed A. Dadamouny , Martin Unterseher, Peter König, Martin Schnittler

University Greifswald, Institute of Botany and Landscape Ecology, Soldmannstr. 15, D-17487 Greifswald, Germany

* Corresponding author. E-mail: [email protected], [email protected]. Fax: +49 3834 86 41 14

Fig. S1. Tree morphology: One M. peregrina tree at W. Zaghra (a), the leaf (b), seedlings in the Botanical Garden Greifswald, Germany (c), flower (d), floral diagram (e), fruits or pods (f), and seeds (g).

69 Appendix III Population performance and conservation of Moringa peregrina

Fig. S2. Threats affect the existence of M. peregrina in Sinai Egypt; drought for several years (a –d), cutting and over-collection (e –f) overgrazing (g).

Fig. S3. Distribution map of Moringa peregrina in all verified regions, using GeoCAT (http://geocat.kew.org/) considering all available coordinates (measured for Egypt’s populations and georeferencing for other coordinates using different distributions maps). The number of points may be underestimation but the extent of occurrence and the area of occupancy are overestimated because the species is confined to some wadis in its wadi beds, slopes, and mountains where water and it is not recorded in the large sandy deserts. A-D the four wadis in Sinai (W. Agala A, W. Feiran B, W. Meir C, and W. Zaghra D), E. worldwide, and F. without Oman and UAE. 70

Appendix IV

Population performance of Moringa peregrina (Forssk.) Fiori (Moringaceae) at Sinai Peninsula, Egypt in the last decades: Consequences for its conservation

∗ Mohamed A. Dadamouny , Martin Unterseher, Peter König, Martin Schnittler

Ernst-Moritz-Arndt-University Greifswald, Institute of Botany and Landscape Ecology, Soldmannstr. 15, D-17487 Greifswald, Germany

* Corresponding author. E-mail: [email protected], [email protected]. Fax: +49 3834 86 41 14

Table S1. Distribution of M. peregrina in fifteen countries (12 are verified) all over the word. This species is the only species in seven countries correspond to its status according to IUCN criteria; Vulnerable VU, Endangered EN, and not evaluated NE.

Country # sp. Recorded species Status Data Sources

1 El-Hadidi et al. (1992), Boulos (1999), Batanouny et al. (1999), Dadamouny Egypt 1 VU 1 Moringa peregrina (2009), Gomaa & Picó (2011), Hassler (2016) EN* This study M. peregrina, stenopetala, Olson (1999), Gomaa and Picó (2011), Lalas et al. (2012), HMT (2013), Sudan 3 NE oleifera Seid (2013). Eritrea 2 M. peregrina, M. rivae2 2VU Edwards (2000), Somali et al. 1984, Gomaa and Picó (2011), HMT (2013)

M. peregrina, M. stenopetala, M. Jahn et al. (1986), Olson (1999a), Edwards (2000), Edwards & Hedberg Ethiopia 5 rivae2, M. ruspoliana, M. 2VU (2000), Vivero et al. (2005), Melesse et al. (2009), Steinitz et al. (2009), longituba Gomaa and Picó (2011), Gebregiorgis et al. (2012), Helen (2014).

Lebrun et al. (1989), Kürschner (1998), HMT (2013), Gomaa & Picó (2011), Djibouti 2 M. peregrina, M. stenopetala NE Seid (2013), Helen (2014) Thulin & Warfa (1987), Kürschner (1998), Oluduro & Aderiye (2007), M. borziana*, M. peregrina, M. Melesse et al. (2009), Steinitz et al. (2009), Gomaa and Picó (2011), Somalia 5 stenopetala, M. pygmaea, M NE Gebregiorgis et al. (2012), Ithaka (2012), Lalas et al. (2012), Padayachee longituba and Baijnath (2012), Bichi (2013). Somali et al. 1984, Al-Kahtani (1995), Miller & Cope (1996), Tsaknis (1998), Saudi 1 M. peregrina, M. oleifera (?)** NE HMT (2013), Osman HE and Abohassan (2012), Alatar AA (2013), Arabia Robiansyah et al. (2014), Mridha (2015), El-Awady et al. (2015) Israel/ Hakham and Ritte (1993), Danin (2004), Steinitz et al. (2009), Shmida et al. 1 Moringa peregrina VU Palestine (2008), Shmida et al. (2013), HMT (2013), Hassler (2016). Engel & Frey (1996), Ejoh et al. (2010), Padayachee and Baijnath (2012), Jordan 1 M. peregrina, M. Oleifera (?) NE HMT (2013), Oran (2014), Majali et al. (2015), Matouq et al. (2015). Syria ? Moringa peregrina NE Padayachee and Baijnath (2012), HMT (2013). Miller & Cope (1996), Wood (1997), Kilian et al. (2004), Jahn (2005), HMT Yemen 1 Moringa peregrina NE (2013), Hassler (2016). Miller & Cope (1996), Jongbloed et al. (2003), Ghazanfar, S.A. (2003), Oman 1 Moringa peregrina NE Marwah et al. (2006), Al-Owaisi et al (2014). United Arab .(Moringa peregrina NE Jongbloed et al. (2003 ۱ Emirates HMT (2013), Talebi et al. (2014), Asghari et al. (2015), Dehshahri et al. Iran ? Moringa peregrina NE (2012)

M. peregrina (?), M. Padayachee and Baijnath (2012), HMT (2013), Anwar et al. (2006), Anwar Pakistan ? 3LR concanensis4, M. oleifera and Bhanger (2003), Anwar et al. (2005), Oluduro & Aderiye (2007). * Endemic to the site. ** Cultivated tree.

71 Appendix IV Population performance and conservation of Moringa peregrina

gerbils (Meriones unguiculatus). British journal of Sources nutrition, 103(11), 1594–1601. Engel, T., Frey, W. (1996). Fuel resources for copper smelting in Alatar, A. A. (2013). Effect of temperature and salinity on germination antiquity in selected in the Edom highlands to the of Achillea fragrantissima and Moringa peregrina from Saudi Wadi Arabah/Jordan. Flora (Jena) 191(1), 29–39. Arabia. African Journal of Biotechnology, 10(17), 3393–3398. Gebregiorgis, F., Negesse, T., Nurfeta, A. (2012). Feed intake and Al-Kahtani, H. A. (1995). Moringa peregrina (Al-Yassar or Al-Ban) utilization in sheep fed graded levels of dried Moringa seeds oil from Northwest Saudi Arabia. J. King Saud (Moringa stenopetala) leaf as a supplement to Rhodes grass Univ., (Agricultural sciences), 7(1), 31–45. hay. Tropical animal and health production, 44(3), 511–517 Al-Owaisi, M., Al-Hadiwi, N., & Khan, S. (2014). GC-MS analysis, Springer. determination of total phenolics, flavonoid content and free radical Ghazanfar, S. A. (2003). Flora of the Sultanate of Oman. Vol. 1. scavenging activities of various crude extracts of Moringa Piperaceae. Scripta Botanica Belgica, 25. peregrina (Forssk.) Fiori leaves. Asian Pac J. Trop. Biomed., 4(12), 964–970. Gomaa, N. H., Picó, F. X. (2011). Seed germination, seedling traits, Anwar, F., & Bhanger, M. I. (2003). Analytical characterization of and seed bank of the tree Moringa peregrina (Moringaceae) in Moringa oleifera seed oil grown in temperate regions of a hyper-arid environment. American journal of botany, 98(6), Pakistan. Journal of Agricultural and Food Chemistry, 51(22), 1024–1030. 6558–6563. Danin, A. (2004). Distribution atlas of plants in the Flora Palaestina Anwar, F., Ashraf, M., & Bhanger, M. I. (2005). Interprovenance area. Israel academy of sciences and humanities. variation in the composition of Moringa oleifera oilseeds from Dehshahri, S., Wink, M., Afsharypuor, M., Asghari, G., Pakistan. Journal of the American Oil Chemists' Society, 82(1), 45– Mohagheghzadeh, A. (2012). Antioxidant activity of 51. methanolic leaf extract of Moringa peregrina (Forssk.) Fiori. Anwar, F., Nahid Zafar, S., & Rashid, U. (2006). Characterization of Research in Pharmaceutical Sciences, 7(2), 111–118. Moringa oleifera seed oil from drought and irrigated regions of Hakham, E., & Ritte, U. (1993). Foraging pressure of the Nubian Punjab, Pakistan. Grasas Y Aceites, 57(2), 160–168. ibex Capra ibex nubiana and its effect on the indigenous Asghari, G., Palizban, A., & Bakhshaei, B. (2015). Quantitative vegetation of the En Gedi Nature Reserve, Israel. Biolog. analysis of the nutritional components in leaves and seeds of the Conservation, 63(1), 9–21. Persian Moringa peregrina (Forssk.) Fiori. Pharmacognosy Res., Hassler, M. (2016). World Plants: Synonymic Checklists of the 7(3), 242–248 doi: 10.4103/0974-8490.157968 Vascular Plants of the World (version Nov 2015). In Roskov Batanouny, K. H., Abou Tabl, S., Shabana, M., & Soliman, F. (1999). Y, Abucay L, Orrell, T., Nicolson D., Kunze, T., Flann., C., Wild medicinal plants in Egypt. With the contribution of: E. Bailly, N., Kirk, P., Bourgoin, T., DeWalt, R.E., Decock W, De Aboutabl, M. Shabana & F. Soliman. With the support of the Swiss Wever A (eds) Species 2000 & ITIS Catalogue of Life. Digital Development Co-operation (SDC). Academy of Scientific Research resource at www.catalogueoflife.org/col. Species 2000: and Technology, Egypt. The World Conservation Union (IUCN), Naturalis, Leiden, the Netherlands. ISSN 2405-8858. (accessed Switzerland, 60–64. 18.2.2016). Boulos, L. (1999). Flora of Egypt: volume 1. Azollaceae- Helen, G. (2014). Evaluation of Analgesic and Anti-inflammatory Oxalidaceae. Cairo: Al Hadara Publishing xv, 419p. activity of Moringa stenopetala Bak.(Moringaceae) in Dadamouny, M. A. (2009). Population Ecology of Moringa peregrina mice (Doctoral dissertation, AAU). growing in Southern Sinai, Egypt. M.Sc. thesis, Environmental HMT Healing Moringa Tree (2013). Web page: Sciences, Bot. Dept., Faculty of Science, Suez Canal University, http://www.healingmoringatree.com/other-moringa- Ismailia, Egypt, p 205. species.html (accessed 24 .12.2015). Edwards, S. (2000). Flora of Ethiopia and Eritrea. Magnoliaceae to Ithaka (2012) Jstor Plant Science. Moringa pygmaea Verdc. Flacourtiaceae. National Herbarium, 2(1). Addis Ababa University. [family Moringaceae]. http://plants.jstor.org/flora/flos000847 (accessed 22.12.2015). Edwards, S., & Hedberg, I. (2000). Flora of Ethiopia. Vol. 2. Part 1, Jahn, S. A., Musnad, H. A., & Burgstaller, H. (1986). The tree that Magnoliaceae to Flacourtiaceae. National Herbarium, Biology purifies water; cultivating multipurpose Moringaceae in the Department, Science Faculty, Addis Ababa University. Sudan. Journal of Unasylva, 38(152), 23–28. Jahn, S. A. A. (2005). How Plant Names Reveal Folk Botanical El-Awady, M. A., Hassan, M. M., Abdel-Hameed, E. S. S., & Gaber, Classification, Trade, Traditional Uses and Routes of A. (2015). Comparison of the Antimicrobial Activities of the Dissemination (I). Studia Asiatica. International Journal for Leaves-crude extracts of Moringa peregrina and Moringa oleifera Asian Studies, (6), 81–126. in Saudi Arabia. Int. J. Curr. Microbiol. App. Sci, 4(12), 1–9. Jongbloed, M., Feulner, G., Böer, B., & Western, A. R. (2003). The El-Hadidi, M. N., Abd El Ghani M. M., Fahmy, A.G. (1992). The Plant comprehensive guide to the wild flowers of the United Arab Red Data Book of Egypt. Woody Perennials. Palm Press and Cairo Emirates. Environmental Research and Wildlife Development University Herbarium. ISBN - 977-5089-04-2. Agency. Ejoh, R. A., Dever, J. T., Mills, J. P., Tanumihardjo, S. A. (2010). Small Kilian, N., Hein, P., & Hubaishan, M. A. (2004). Further notes on quantities of carotenoid-rich tropical green leafy the flora of the southern coastal mountains of Yemen. indigenous to Africa maintain vitamin A status in Mongolian Willdenowia, Bd., 34(1), 159–182 doi:10.3372/wi34.34114. 72

Population performance and conservation of Moringa peregrina Appendix IV

Lalas, S., Gortzi O., Athanasiadis V., Tsaknis J., Chinou, I. (2012). Padayachee, B., & Baijnath, H. (2012). An overview of the medicinal Determination of antimicrobial activity and resistance to importance of Moringaceae. Journal of Medicinal Plants oxidation of Moringa peregrina seed oil. Mol., 17, 2330–2334. Research, 6(48), 5831–5839. Lebrun, J. P., Audru, J., & César, J. (1989). Catalogue des plants Salaheldeen, M., Aroua, M. K., Mariod, A. A., Cheng, S. F., vasculaires de la Republique de Djibouti. Etudes et Syntheses de Abdelrahman, M. A., & Atabani, A. E. (2015). Physicochemical L'IEMVT, 34. characterization and thermal behavior of biodiesel and biodiesel– diesel blends derived from crude Moringa peregrina seed Majali, I. S., Oran, S.A., Khaled, M. K., Qaralleh, H., Rayyan W. A., oil. Energy Conversion and Management, 92, 535–542. Althunibat, O. Y. (2015). Assessment of the antibacterial effects Shmida, A., Cohen, O., & Cohen, A. (1988). Moringa: A Tropical of Moringa peregrina extracts. African Journal of Microbiology Tree in the Golan. Israel Land Nat, 14(1), 219–222. Research, 9(51), 2410–2414. Shmida, A., Pollak, G., & Fragman-Sapir, O. (2013). The Red Data Marwah, R. G., Fatope, M. O., Al Mahrooqi, R, Varma, G. B., Al Book: Endangered Plants of Israel. Israel Nature and Parks Abadi, H., & Al-Burtamani, S. K. S. (2006). Antioxidant capacity Authority. Web site: of some edible and wound healing plants in Oman. Food http://redlist.parks.org.il/taxa/Moringa%20peregrina/ (accessed chemistry, 101(2), 465–470 Sep. 2015). doi:10.1016/j.foodchem.2006.02.001 Seid, M. A. (2013). Medicinal and Dietary Role of Moringa Matouq, M., Jildeh, N., Qtaishat, M., Hindiyeh, M., Al Syouf, M. Q. stenopetala (Bak. f.) Cuf. in South Ethiopia: A Review. African (2015). The adsorption kinetics and modeling for heavy metals Journal of Agricultural Science and Technology (AJAST), 1(1), removal from wastewater by Moringa pods. Journal of 1–6. Environmental Chemical Engineering, 3(2), 775–784. Somali, M. A., Bajneid M. A., & Al-Fhaimani S. S. (1984). Chemical Melesse, A., Bulang, M., Kluth, H. (2009). Evaluating the nutritive composition and characteristics of Moringa peregrina seeds and values and in vitro degradability characteristics of leaves, seeds seeds oil. Journal of the American Oil Chemists' Society, 61(1), and seedpods from Moringa stenopetala. Journal of the Science 85–86. of Food and Agriculture, 89(2), 281–287. Steinitz, B., Tabib, Y., Gaba, V., Gefen, T., & Vaknin, Y. (2009). Miller, A. G., & Cope, T. A. (1996). Flora of the Arabian Peninsula Vegetative micro-cloning to sustain biodiversity of threatened and Socotra (Vol. 1). Edinburgh University Press. Moringa species. In Vitro Cellular & Developmental Biology- Mridha, M. A. U. (2015). Prospects of Moringa Cultivation in Saudi Plant, 45(1), 65–71. Arabia. J. Appl. Environ. Biol. Sci., 5(3), 39–46. Tsaknis, J. (1998). Characterisation of Moringa peregrina Arabia Robiansyah, I., Hajar, A.S., Al-kordy M. A., & Ramadan, A. (2014). seed oil, Journal of Grasas y Aceites, 49(2), 170–176. Current Status of Economically Important Plant Moringa Talebi, K. S., Sajedi, T., & Pourhashemi, M. (2014). Forests of Iran: peregrina (Forrsk.) Fiori in Saudi Arabia: A Review. A Treasure from the Past, a Hope for the Future. Chapter 4: International Journal of Theoretical & Applied Sciences, 6(1), Saharo-Sindian Region, pp 115–141 Plant and vegetation doi 79–86. 10.1007/978-94-007-7371-4_4 Olson, M. (1999). The home page of the plant family Moringaceae http://link.springer.com/chapter/10.1007/978-94-007-7371-4_4 http://www.mobot.org/gradstudents/olson/moringahome.html (accessed Dec. 2015). (accessed 15.8.2015). Vivero, J. L., Demissew, S., & Kelbessa, E. (2005). The Red List of Oluduro, A. O., & Aderiye, B. I. (2007). Efficacy of Moringa oleifera Endemic Trees Et Shrubs of Ethiopia and Eritrea. Fauna et Flora seed extract on the microflora of surface and underground International, UK. water. J. Plant Sci., 2,453–458. https://portals.iucn.org/library/sites/library/files/documents/RD- Oran, S. A. (2014). The status of medicinal plants in Jordan. Journal 61-001.pdf of Agricultural Science and Technology. A 4, 461–467. Wood, J. R. I. (1997). A handbook of the Yemen flora. Kew: Royal Osman, H. E. & Abohassan A. A. (2012). Morphological and Botanic Gardens, Kew vi, 434p, 40. analytical characterization of Moringa peregrina populations in Western Saudi Arabia. Inter J Theor Appl Sci., 4(2), 174–184.

73 Appendix IV Population performance and conservation of Moringa peregrina

Table S2 The description of the morphological and anatomical charterers of M. peregrina and the threats affect its existence.

Description It is deciduous, perennial tree (Fig. S1a). Its main root is thick. Its branches and stem are brittle with corky bark, (Fig. 4a and b). Its trunk is erect, terete, branched, and branches are divaricated or ascending, slender, forming avoid or abavoid crown, green-glaucaus (El-Hadidi et al., 1991). M. peregrina is a deciduous tree since its leaves are shed. Tree Morphology The leaves are feathery (Fig. S1b), pale green, compound tri-pinnate leaves that can reach 30-60 cm length (Boulos, 1999). The leaflets are typical between 1.3 and 2.5 cm long, develop only after pronounced rainfalls and will be shed during prolonged periods of drought. Also, the leaflets are simple, petiolate, and glabrous on both surfaces, the blade is ovate- oblanceolate, margin entire, and apex obtuse, sometimes mucronate (El-Hadidi et al. 1991). The tree endures by its green assimilating young twigs, and this growth form wasn't common in the floras of sub-Saharian Africa and the Arab Peninsula. Seedlings are shown in Fig. S1c. Flowering system Flowers of M. peregrina are stalked on panicles (Figure S1.dand e), fragrant, with white, creamy, pinkish to pale, 2.5 cm in diameter (Batanouny et al. 1999). The five sepals are oblong-lanceolate, acuminate, whitish, with five white-pinkish to pale petals and five yellowish stamens (El-Hadidi et al. 1991; Olson 2003). Although Moringa is bisexual, Abd El-Wahab (1995) and Dadamouny (2009) stated that Moringa produces a rather high number of flowers in comparison to the comparatively low number of pods. Fruits The fruits or pods of M. peregrina are pendulous ridged, brown, triangular. The ribbed pods reach from 15 to 30 cm length (Fig. S1h) and split at maturity lengthwise into 3 parts exposing about 20-25 dark brown seeds embedded in the pith (Figure S1g,). Seeds description Seeds have no dormancy, and up to 90% of a cohort germinate (Fig. S1c), which is in sharp contrast to and germination the low establishment rate of the Sinai populations, where old trees are dominated (Dadamouny 2009; Dadamouny et al. 2012). The seeds have a bitter-sweet nauseous taste and are rich in oil (ben-oil) (Täckholm, 1974). No prior treatments for germination (only soaking the seeds in water for 24 hours).

Phenology M. peregrina becomes flowering in March and April. Unripe fruiting period occurs from April to June. Ripe pods can be collected at the end of June and July, occasionally in September and October especially those grown on the Red Sea Mountains (Zahran and Willis 2009) stated that pods ripen in October. The mature tree produces about 150 pods of 10 to 15 seeds per each (1700 seeds/1 kg.). Finally, at the Sinai (our study area), December and January are the dry periods although the M. peregrina trees still having some green petioles (Dadamouny 2009). Threats The species is a desert tree (tropical to sub-tropical) producing seeds and not reproducing by vegetative fragmentation. The seeds are woody, rich in proteins, need dry conditions storage, not remain in the soil seed bank, and its viability can be reduced to 50 % after 6 months to 1 year. Low establishment and absence of recruitment. Overgrazed (palatable, not toxic, without spines or thorns). Over-collected by humans (medicinal parts, the target seeds, rich in proteins, vitamins, minerals, and anti-inflammatory compounds, wood for fuel). The tree is under the stress due to: climate change (an increase in the area aridity due to the increase of the temperature and rareness of the rainfall), in addition, the overuses by Human.

* Sources are listed under Table S1 Sources Olson, M. E. (2003). Ontogenetic origins of floral bilateral symmetry in Dadamouny, M. A., Zaghloul, M. S., & Moustafa, A. A. (2012). Impact Moringaceae (Brassicales). American J of Botany, 90(1), 49–71. of Improved Soil Properties on Establishment of Moringa Tackholm, V., & Drar, M. (1956). Students' flora of Egypt. Cairo: peregrina Seedlings and a Trial to Decrease its Mortality Rate. Anglo- Egyptian Bookshop. Egyptian J. Botany, 52(1), 83–98. Zahran, M. A., & Willis, A. J. (2009). The vegetation of Egypt, 2nd Ed., London: Chapman and Hall, p. 437.

74 Population performance and conservation of Moringa peregrina Appendix IV

Table S3. The status of M. peregrina according to the IUCN criteria 2016. The column of local status is referred to the evaluation of the censused populations. Local Status Criteria Worldwide Evaluation (Estimated Data) Local Evaluation (Measured for Egypt) 2016 Population Moringa peregrina is native to regions with tropical or The populations are fragmented into four wadis in Sinai (W. Vulnerable Size (A1) subtropical climates. Its Populations have an observed and Agala WA, W. Feiran WF, W. Meir WM, W. Zaghra WZ) & a estimated reduction in population size, inferred, or suspected in population confined to the Red Sea Mountains. the last three decades where the causes of the reduction (climate, Reduction in population size at Sinai from 448 (in 2000) to 390 human pressure) are clearly reversible. (2015), without any establishment for any new individual during The percentage of population reduction could be more than 10 the last three decades. % in the last 30-40 years as a result of global climate change and As the tree is a profitable plant, it suffers from over-collection the over-use of the tree as one of the main medicinal plants in for medicinal uses & as well over-grazing. A reduction in Egypt the Arabian Peninsula. The percentage may increase at Saudi is estimated to be 15 % in 10 years. Arabia and Yemen, as the tree populations, there are suffering The estimated age for the youngest tree (in 2015) was 27 years also from over-collections (1 Kg of seeds exceeds 1000 Saudi at Sinai, Egypt, & therefore M. peregrina tree has a great Riyal (≈ 300 USD). problem in recruitment & establishment. More M. peregrina is recorded in Saudi Arabia KSA, Egypt, Palestine, Based on the tree performance, current recruitment for Sinai Endangered Population Israel, Oman, Yemen, United Arab of Emirates UAE, Sudan, populations, decrease of tree tolerability & adaptation to the reduction in Ethiopia, Somalia, Eritrea, & Djibouti. environmental constraints, the remaining M. peregrina Future (A3) Its populations are projecting to a reduction at least in the first populations are expected to decrease to less than 50 % in the seven countries which have only this species of Moringaceae. next 50 years. Continued overuse & climate severity could affect the existence Without conservation efforts, this species will completely of the tree in these countries in the near future (70 to 100 years), disappear in the next 70 to 100 years. especially in Egypt, Palestine, KSA, & Israel. Extent of 3,824,965 km2 the whole area including Oman and UAE. 1,020.3 km2. (It is overestimated because the species is not Endangered occurrence 1,608,527 km2 the Red Sea regions with its gulfs (area without recorded in the area between the four wadis of Sinai). 2 2 2 (EOO) (B1) Oman & UAE). 2.201 km in WM, 0.155 km in WA, 0.219 in WF, & 3.750 km 2,239,778 km2 Arabian Peninsula (KSA, Yemen, Oman, & (with one outlier tree) & only 0.444 km2 (without outlier). UAE), 84.673 km2 Israel + Palestine. (i.e. EOO is lower than 5,000 km2). 223,277 km2 KSA, 30,497 km2 Oman, 6,441 km2 Yemen. Area of 224 km2 worldwide. 36 km2 Sinai including the area between the wadis but for the occupancy 184 km2 the Red Sea regions (area without Oman & UAE). total wadis only is 18.56 km2 (WA = 0.540 km2, WF = 0.270 2 2 2 2 (AOO) (B2) 100 km Arabian Peninsula (KSA, Yemen, Oman, & UAE). km , WM = 16 km , & WZ = 1.75 km (with one outlier tree, 1.5 120 km2 Arabian Peninsula, Palestine, & Israel. km2 without the outliers). 40 km2 KSA, 36 km2 Oman, 20 km2 Yemen. (i.e. less than 40 km2 Sinai & the Red Sea Mountains (including 20 km2 Israel + Palestine, 24 km2 Somalia. the water area of the Red Sea, and the area between these five 44 km2 Sudan + Eritrea + Ethiopia + Djibouti. locations which are free from any tree). Geographic Remaining the decline within in the above EOO/AOO which # of locations = 5. Fragmented populations. range (B1) & could lead to more population fragmentation, limiting its Danin (1999) surveyed many M. peregrina trees on the west- (B2) distribution to a small niche (the distribution edges in Oman or facing escarpments of Serbal Mountain (in W. Feiran), but Egypt) subjected to extremes in environmental conditions these trees were completely dried & disappeared since 2006. (aridity; increasing temperature & rainfall rareness) & Also, 11 Trees of two sites in WZ were completely uprooted decrease/intrusion of ground water. after the flood of 2010. M. peregrina Trees of the sole site in The tree is confined to marginal wadis & the AOO/EOO are Gebel Elba (7 trees) & Gebel Abou-Dukhan and Gebel El- overestimated by GeoCAT because the there is no any record Faryid (surveyed by Kassas & Zahran, 1962) are also for M. peregrina trees in the large sandy desert of the Arabian disappeared. Not recorded in high mountains area like St. Peninsula. Catherine Mountains (temp. is too low) & not house the large wadis which are subjected to flash floods (e.g W. Watir). Extreme fluctuations in both EOO & AOO due to decrease The populations are fragmented in four wadis (sub- number of locations, subpopulations, & number of mature populations), the trees in the largest wadi occupies 16 km2. individuals & low chance of the establishing of the new The dominance of old trees & absence of recruitment (100 % sapling. mature individuals).

Small A number of estimated mature individuals of M. peregrina 420 trees in declining year after year (i.e. < 2,500). Highest Endangered population trees is not more than 10,000 within three main regions, Oman, recorded number of the largest wadi is WM = 242 trees, WZ size decline Yemen, Egypt, & Saudi Arabia. In the other countries (mainly = 71 trees, WF = 45 trees, & WA = 35 trees. (C) in the Horn of Africa), the dominant species is M. oleifera. The youngest tree (36 years in WM, 32 in WA, 27 in WF, and 30 in WZ.

C1 For the estimated M. peregrina trees, a continuing decline of at 10% declined in the last 10 years & it might be 100 % in the Vulnerable least 25 % within the next 70–100 years. next 70–100 years (in case of continuing a bad conservation). C2 M. peregrina has the highest number of mature individuals in The highest recorded number of the largest subpopulation = Endangered KSA, Egypt, & Oman. The number of these individuals in each 242 trees (i.e. the number of mature individuals < 250). subpopulation is lower than 2500. More than 95 % of mature individuals in one subpopulation. 95–100% Restricted So far, no data on the total of M. peregrina worldwide, but in In Egypt, 420 trees (390 in Sinai & 30 trees recorded on the Vulnerable population most sites, it has very small populations, the most are old trees. Red Sea Mountains & W. Esel [i.e. with the non-counted trees (D) on the Red Sea mountains, the total individuals < 1000] Very small The tree is restricted to the mountainous area around the Red All populations confined to 19 km2 (AOO four population are population Sea, gulfs, and some wadis in fragmented small populations at in less than 3 km2. Based on the decline in population size (D2) Arabian Peninsula, Sinai Peninsula, and occasional sites at the within the last three decades, unless the conservation efforts Horn of Africa. In all countries, AOO is lower than 40 km2 in the threatened populations could disappear in the near future each country. from its natural habitats. Extinction in No enough data to evaluate the probability of extinction in the ND NE the wild (E) wild. 75 Appendix IV Population performance and conservation of Moringa peregrina

Table S4. The status of M. peregrina according to the IUCN criteria; first by El-Hadidi et al. (1992) listed in the National Red List of Egypt and the updated current status.

Status 1992 (IUCN 1992) Status 2015 (This study) Criteria IUCN pre-1994 IUCN ver. 12 2016 Systematic Kingdom: Plantae Kingdom: Plantae Phylum: Spermatophyta Phylum: Tracheophyta Subdivision: Angiospermae, Dicotyledoneae Class: Magnoliopsida Class: Rosopsida Order: Capparales Order: Brassicales Family: Moringaceae Family: Moringaceae Genus: Moringa Genus: Moringa Species: peregrina Species: M. peregrina Species Authority: (Forssk.) Fiori Species Authority: (Forssk.) Fiori Distribution M. peregrina is native to the area around the Arabian Peninsula and Recorded from Ethiopia, Somalia and the Sudan eastwards to the Red Sea with its Gulfs. The tree is recorded from NE Africa Arabia. Also known from Palestine and Jordan. A rare species (Egypt, Sudan, Ethiopia, Somalia, Eritrea, and Djibouti) and SW in Egypt and is recorded from the Eastern Desert, Red Sea Asia (Israel, Jordan, W and SW Saudi Arabia, Yemen, Oman, UAE). coast, Gebel Elba mountains and Mountainous Sinai. In Egypt: South Sinai and Red Sea Mountains (Gebel Shindodai, Samiuki, Nugrus, and Shayeb). Habitat Micro-phanerophyte which grows in crevices and slopes. Mountains foothills, crevices and slopes, gorges and wadi beds. Threats Once its value as a cash crop goes out of fashion it may be Over-collection for its pods and seeds for different uses, overcutting, burnt by the Ma'aza Bedouins as fuel. overgrazing especially in the dry season, and consequently, low establishment and miss-recruitment. Justification Due to its requirement for high water, it confined to the feet of As a result of climate change and the rareness of rainfall, the trees coastal hills, which explains its limited distribution. turn into dry which encourages the Bedouins to log. In dry seasons, Moringa is one of the few trees withstanding the drought and keeping its green branches for long times. At this time, it became (the main available resources) a victim of overgrazing. Study area Sinai (Feiran Oasis), Red Sea Mountains at the Eastern Desert. Sinai populations (W. Feiran, W. Agala, W. Meir, W. Zaghra). EOO/AOO is 1,020/36 km2 by GeoCAT but in fragmented populations. The area between these occasional sites is free of trees. Conservation Gebel Elba protected area. Also cultivated in El Orman garden 200 trees were cultivated in 2001 by Dadamouny MA at the Measures as an ornamental tree. Botanical Garden, Faculty of Science, Ismailia, Egypt. But up to 2015, only 17 trees remained. Four old trees were fenced at W. Zaghra in 2004 by St. Catherine Protectorate Authority. The enclosure was removed by Bedouins in 2011. Recommended The populations in Wadi Bali and Gebel El Shayeb should be The old populations of W. Zaghra (north-east St. Catherine) and the Measures declared wildlife sanctuaries. young populations of W. Meir (north-east El-Tor) should be fenced and two or three Guards of Bedouins employed to protect these areas. Cultivation of the collected seeds samples inside the enclosures at its natural habits (preferably the foothills of the Sinai and the Red Sea mountains) and ex-situ conservation (e.g. at Wadi El-Gemal-Hamata National Park along the Red Sea mountains). Assessors Hadidi MN; Abd El Ghani MM; Fahmy AG Dadamouny MA; Unterseher M; König P; Schnittler M Year Assessed 1992 2006–2015 Regional Vulnerable (VU). Endangered (EN). Status

76 Population performance and conservation of Moringa peregrina Appendix IV

Table S5. Descriptive statistics (mean, standard deviation, standard error, minimum, and maximum) of the tree fitness parameters for Moringa peregrina trees growing in north-facing and south-facing.

North-facing South-facing Pods Pods Seeds Seeds Pods Pods Seeds Seeds pod_d Se_d pod_d Se_d 2006 2014 2006 2014 2006 14 2006 2014 N 189 189 189 189 189 189 204 204 204 204 204 204 13 19.6 133.6 189.8 7.1 56.2 12.0 22.8 127.7 227.4 10.8 99.7 Mean St.D. 22.3 22.3 29.0 237.2 20.4 197.4 1.4 20.4 35.8 223.4 28.0 279.7

St.E. 1.6 2.1 17.3 21.2 1.5 14.4 20.4 2.5 15.6 26 2.0 19.5819 0 Min. 0 0 0 0 -38 -500 0 0 0 -45 -496.6 Max. 117 173 1456 1863 120 1101 110 340 1377 3196 230 2162

Table S6. Matrix of Pearson correlation coefficients (r) showing correlations between tree elevation above wadi surface, vitality, and seed set (number of pods, number of seeds and the differences) for M. peregrina trees growing in north-facing (N=189) and south-facing (N=204).

North-facing South-facing Elev Vit Pods Pods Seeds Seeds Seeds_ Elev. Vit. Pods Pods Seeds Seeds Pods_d Pods_d Seeds_d a.w.s 2006 2006 2014 2006 2014 d a.w.s 2006 2006 2014 2006 2014 Elev. a.w.s. 1.00 1.00 Vit. 2006 –0.25 1.00 –0.22 1.00 Vit. 2014 –0.41 0.56 –0.32 0.48 Pods 2006 –0.19 –0.02 1.00 –0.24 0.17 1.00 Pods 2014 –0.25 –0.01 0.71 1.00 –0.31 0.09 0.63 1.00 Seeds 2006 –0.22 0.00 0.97 0.73 1.00 –0.23 0.17 0.97 0.59 1.00 Seeds 2014 –0.25 0.00 0.71 0.97 0.74 1.00 –0.30 0.10 0.69 0.97 0.67 1.00 Pods_d –0.14 0.02 –0.08 0.64 –0.02 0.61 1.00 –0.22 -0.01 0.07 0.82 0.05 0.74 1.00 Seeds_d –0.11 0.00 –0.13 0.56 –0.11 0.59 0.93 1.00 –0.22 0.00 0.16 0.84 0.11 0.81 0.96 1.00

77

3.3 Demography and recruitment in populations of the threatened tree Moringa peregrina (Moringaceae) at Sinai Peninsula, Egypt

Mohamed A. Dadamouny and Martin Schnittler Ernst-Moritz-Arndt University of Greifswald, Institute of Botany and Landscape Ecology, Soldmannstr. 15, D-17487 Greifswald, Fax +49 03834 420 4114 *Email: [email protected]

Key message Adapted to arid habitats, Moringa peregrina is one of the few subtropical trees displaying annual growth rings that allow age estimation. The relationship between age and tree radius could be used to determine the age structure of four populations, which showed prolonged periods of missing recruitment. Abstract Based on age determination of wood samples, the age structure of four remnant populations of Moringa peregrina at Sinai Peninsula, Egypt was determined using three alternative methods. For the endangered tree species any destructive sampling had to be avoided, therefore age estimation was based on counts of cross sections of dead and broken branches, trunks of recently dead and uprooted trees, and a relation between age and diameter was established. Since many trees grew multiple trunks due to frequent damage, tree age was estimated based on a) the mean of the radius measurements for all trunks of a tree, b) the radius of a virtual trunk calculated from summing up the cross area of all trunks, and c) the radius of the largest trunk. Possible errors like differences in bark thickness and missing or false rings were kept to a minimum. Age and size structure was assessed for 393 trees in 2007 and 2015, and data on phenology and vitality were additionally recorded. A life table and survivorship curves were calculated, and age-dependent recruitment was estimated. The age distribution for the recorded Moringa trees indicated highly fluctuating, but decreasing recruitment rates, showing numerous failure years that did not add any new individuals to the populations. By coincidence the youngest and the oldest tree were single- stemmed, resulting in age estimates of 27 and 386 years, respectively. For multistemmed trees, age estimates according to the three methods varied but did consistently show a trend towards decreasing recruitment. With the severe decline in population size recorded over the last four decades, M. peregrina should be updated on the IUCN Red List of Threatened Species, requiring an urgent conservation program.

Keywords: Age determination. Dendroclimatology. Life table. Mis-recruitment index. Survivorship curve. 1. Introduction M. peregrina is native to Arabian Peninsula, and possibly the horn of Africa and as far north as the Dead Sea. It is found from Ethiopia and Somalia, northwards to Sudan and eastwards to Eritrea and Djibouti (Olson 2002). It is also recorded in Asia at Saudi Arabia, Palestine and Israel, Syria, Jordan, Oman, and Yemen (Dadamouny 2009, et al. 2016). In Egypt, it occurs at the margin of its range at Sinai and the Red Sea mountains. The remaining populations of M. peregrina (Appendix 1 Fig. S1a) as one of the few desert plants at the southern Sinai were long suspected to be in steady decline (Dadamouny 2009). This was due to human overuse and climate changes with a trend towards warmer and drier years or higher fluctuations in rainfalls (Dadamouny and Schnittler, 2016). This pressure caused a threat status for the remaining populations, in addition to the aridity of the area that led to poor survival and recruitment. The threats seem to be accelerated by a low establishment 78

Demography and recruitment of Moringa peregrina Publications

rate and poor recruitment. Therefore, a long-term study on its population dynamics was started (Dadamouny 2009; Dadamouny et al. 2012). The current study will report results on the count of annual growth rings (Fritts et al. 1989; Schweingruber 1988; Fritts et al. 1991) from dead and broken branches and trunks to estimate the age structure of the populations, and finally, reveal the recruitment patterns for the four threatened Moringa populations at Sinai over time (Dadamouny et al. 2016). Initial mapping and complete censuses for these populations were carried out in 2006, 2007, 2015 and 2016, and these results were already used to assess habitat quality and to derive recommendations for a conservation program of this threatened tree species (Dadamouny 2009, et al. 2012).

Studying the demography of trees poses various approaches and obstacles (Harper and White 1974). Most commonly used are increment cores method (Ferguson 1970) which is cheaper than radiocarbon dating. Although the tree coring was well established for trees from temperate zones (Schweingruber 1988), it may harm the trees and not suitable for the lightweight wood trees. Already Coster (1927) described growth rings in tropical trees, however, it was not clear if these represent true annual growth rings or not (e.g. Ogden 1981, Lieberman et al. 1985). According to many recent studies (e.g. Worbes 1999, et al. 2003; Schöngart et al. 2002; Chowdhury et al. 2008, Rozendaal and Zuidema 2011, Schweingruber 2012), such annual rings are formed in tropical or subtropical trees if they experience cambial dormancy during a certain period of the year caused by environmental conditions. These conditions might be unfavourable for growth nevertheless these are part of the annual climatic and phenological changes (like dry seasons, flooding periods in floodplain forests, or fluctuations in salinity (e.g. mangrove forests, Rozendaal and Zuidema 2011). Even in forests, such studies succeeded in the areas where a short dry period triggers annual ring formation (e.g. Dünisch et al. 2003; Fichtler et al. 2003). However, M. peregrina trees are obtaining large quantities of water storage tissue that made it anatomically less important than many other woody trees (Olson and Carlquist, 2001). This spurred first the question if observable annual growth rings can be used to age this trees (Dadamouny 2009). Estimates from dead and broken branches and trunks, which could be used without inflicting further harm to the threatened populations were used to determine the current demographic status of M. peregrina populations. As one measurement is not enough to assess if the population changed positively or negatively, two repeated tree measurements were used to calculate the increment in that period, test if the M. peregrina tree produces annual growth ring or not, and if data derived from cross cuts allow a reliable age estimation. The data were used to construct a tree life table and to estimate age-dependent tree mortality and recruitment. Recruitment is the capability of adding individuals to the population; in a wide sense, it includes as well survival when trees move from one size class to another and finally reproduce. Similarly, recruitment can be monitored by counting all saplings or young trees that overgrow a certain threshold value of height or dbh (Lexerød and Eid 2005, de Avila et al. 2017). In this study, we identified the recruitment of M. peregrina by ageing the trees and identifying their year of germination from seed. 2. Material and Methods 2.1. Study area Repeated surveys at the Sinai Peninsula (2007, 2015, 2016) identified relict populations of the trees with a total of 393 Moringa trees in four wadis at the Southern Sinai; Agala (WA, 35 trees), Feiran (WF, 45 trees), 79 Publications Demography and recruitment of Moringa peregrina

Meir (WM, 242 trees) and Zaghra (WZ, 71 trees). Further details about the study area are provided by Dadamouny (2009); Zaghloul et al. (2012) and Dadamouny et al. (2012, 2016). 2.2. Moringa trees measuring and sampling Size parameters of the trees (height and crown area) were measured in 2007 and 2015 (see Dadamouny et al. 2016). Tree vitality was assessed according to a scale ranging from 1 to 5 (avoiding the time of leaf shedding due to phenology). The scale is described by Dadamouny et al. (2012 and 2016). In 2006, all trees were surveyed and mapped using GPS coordinates (UTM) and permanently labelled for long-term monitoring. As the base figure for age estimation, we measured the circumference at the ground (CAG) in 2007 and 2015 (Fig. S1b). For the analysis of growth rings, we collected a total of 117 cross sections (without repeats) from trunks of recently dead and uprooted trees, as well as dead and broken branches (Zaghloul et al. 2012). These cross sections were 74 from W. Meir, 13 from W. Agala, also 13 from W. Feiran, 17 from W. Zaghra. In spite of causing at least some harm to the trees, the trunk coring was found to be not suitable because the wood structure of M. peregrina is soft and easily destroyed with a small increment borer (Fig. S1c) which inflicts only minor injury to the tree (Zoltai 1975). A larger borer was less detrimental to the wood structure but caused much more damage. We thus applied non-destructive sampling by obtaining cross-sections of dead trunks or broken branches (Zaghloul et al. 2012), aged these and established an age – diameter relationship. This sampling technique was applied by other authors (e.g. Lawerence 1950; Roque et al. 2015, Jacquin et al. 2017). Cross sections were sanded to allow better resolution, the mean radius was determined by eight measurements in 45° distance, and the annual growth rings were counted under a dissecting scope in three directions for each cross section. We calculated the mean of counting for all wood repeats collected for each tree. False and missing rings were identified by comparing different cross sections of similar diameter (wood are shown in Fig. S2). The bark thickness was recorded together with the eight radii measurements, differentiating between upper bark (b1, dead, corky tissue) and the partially alive lower bark (b2). 2.3. Age calculations and error handling Counting annual growth rings are the best method to determine the exact age of trees (Worbes et al. 2003; Santos et al. 2015). As all trees are grown under the same conditions (rains, aridity, wadi resources, etc) and to avoid the damage to the remaining M. peregrina trees, we estimated the tree age by establishing a relationship between trunk diameter and age (Boukili et al. 2017; Stovall et al. 2017) based on ring counts and diameter measurements of 117 wood samples (Batcheler 1985). First, we established a relationship between wood radius and bark thickness to reduce the error in age estimation caused by the bark (Fig. S2). This enabled us to derive a relationship between the measured radius, excluding bark thickness, and the counted growth rings. The obtained two equations are used to estimate the bark thickness and the tree age of 393 living trees recorded in the two censuses of 2007 and 2015. The most severe problem was posed by the high proportion of multistemmed trees, which were destroyed by flash floods and resprouted later (Figs S1, S3). To account for this uncertainty, we calculated the tree radius based on three proposed methods. The first method is the mean of radii, in which we calculated the radius and the bark thickness of each stem separately from its circumference at ground CAG (C = 2πr). The mean value of all radii was used for age estimation. The second method is virtual single radius, which is calculated from the cross-sectional area of each stem at ground (Dean 2003; Sperry and Saliendra 1994) and by summing up these figures, we calculated a virtual single radius (Dadamouny et al. 2016). Finally, 80 Demography and recruitment of Moringa peregrina Publications

the third method is the radius of the largest stem to estimate the age of the largest/oldest stem only. For all three methods, we applied the correction for bark thickness as outlined above. These calculations were carried out for both censuses (2007 and 2015). 2.4. Data treatment The methods A (mean of all radii) and B (virtual single radius) tend to overestimate age for multi-stemmed trees, while method C (largest/oldest stem) may underestimate it. Only for single-stemmed the outcome of all three methods is identical. Additional sources of error are provided by the correction for bark thickness and false or missing rings. We tried to detect false rings by matching different crosscuts from trees with similar age (cuts from the same multistemmed tree or from the nearest tree of the respective age in the same wadi) to see if there was any false ring formed during the last three to five decades (Bouriaud et al. 2005; Copenheaver et al. 2006), and such rings were excluded from counts. All data were treated using the statistical program R (Available online: http://dx.doi.org/10.17632/z23sx6vjc3.1) and NCSS 10 (2015) and QI Macros 2015. The number of trees (Nt) established in each year or in each age class was determined to reconstruct tree recruitment over time. To reduce the effect of age over- / underestimation, we used age classes of 20 years to visualize tree recruitment. 2.5. Estimating parameters for a static life table and plotting the survivorship curve Estimated M. peregrina tree ages were used to determine the age structure and distribution of the populations and build a life table. We used this distribution as a predictive tool to determine if the populations in Sinai are in a growing or a declining status (Dadamouny 2009). We also used multiple cohorts or snapshots in the tree life cycle to assess the survival probability for each age class in the calculated life table (Cox 1972, Sharitz and McCormick 1973, Silvertown 1982, Bengon et al. 1996, Molles and Manuel, 2002). For this table, the following parameters are used: x = age entered by the time of the census, Nx = a number of individuals living in age class x with ax = a number of individuals that survive to age x. Survivorship lx was calculated as the

proportion of original cohort surviving to age x (individuals at each age interval ax divided by the total number of individuals surviving until to the time of assessment a0) and used to construct a Kaplan-Meier or survivorship curve (equation 1). This determines Lx as the average proportion alive at that age (Equation 2) and Tx as the total

number of living individuals at age class x and beyond (Equation 3). Therefore, ex will be the probability of living 'x' number of years beyond a given age (Equation 4), and dx the number of individuals that die during that time interval (Equation 5). Finally, the age-specific mortality rate qx can be defined as the proportion of individuals entering age x that die during age x (Equation 6). These parameters can be derived from survivorship

values lx as follows:

lx = ax/a0 (1) Lx = (lx + lx+1)/2 (2)

Tx = ∑ Lx – x-1 (e.g. T3 = L3 + L4 + L5 + ....+ Lmax) (3) ex = Tx / lx (4)

dx = lx – lx+1 (5) qx = dx/ax (6)

2.6. Mis-Recruitment Index and Recruitment Failure Based on the number of the established trees, we could count the failure years in which the population could not add or establish new individuals. We calculated an index describing the proportion of recruitment failure within a population (Mis-Recruitment Index). MRI = [Nd/PA] where Nd is the number of failure years (years where no tree was established). One can calculate Nd from the difference in the tree age and the age of the tree 81 Publications Demography and recruitment of Moringa peregrina

beyond this age at the time of assessment. PA is the population age at the time of the assessment (the oldest alive individual). For any population, the total of the Mis-Recruitment index equals one. The percentage of recruitment failure (RF %) = [MRI × 100] indicates the proportion of years where the population did not grow. The highest failure intervals were predicted using Weibull analyses, where, the Weibull distribution gives a distribution for which the failure rate is proportional to a power of time (population age). It indicates if the failure rate decreases or increases over time. It depends on the mortality rate especially for saplings or young trees. From Weibull distribution, we expected the intervals in which the population has a high failure rate (fail to add new individuals or decrease in population size). 3. Results 3.1. Field census A total of 426 trees were surveyed in 2006; of these, only 393 trees (92%) could be relocated at the 2015 census. Fig. 1 shows the spatial distribution of trees. Most of the trees were recorded to the slopes and foothills of the mountains. The highest tree vitality based on a scale 1–5 was recorded for the mapped trees growing in wadis, but these were most often multistemmed, indicating to be damaged or uprooted either by human impact or flash floods.

Fig. 1. Satellite images (Google Earth) showing the geographical distribution of all recorded M. peregrina trees (a) worldwide: labelled green dots, (b) study area in Egypt, (c) recorded wadis in Sinai, and the recorded 393 trees in 2015 at the four investigated wadis: W. Agala (d), W. Feiran (e), W. Meir (f), and W. Zaghra (g). Each tree is indicated by a site number (S, 41 in total) and a tree number (T).

3.2. The relationship between the tree radius and annual growth rings About 60–80 % of the trees are multi-stemmed (Appendix 1 Fig. S3; 22 of 35 at W. Agala. 28 of 45 at W. Feiran, 168 of 242 at W. Meir, and 58 of 71 at W. Zaghara), which hampers estimation of the trunk diameter. However, both the youngest and oldest tree are single-stemmed. Bark thickness is weakly correlated with total trunk radius (R2=0.253, Fig. 2A), but the correlation of the corrected radius with the number of growth rings (age) is much better (R2=0.592, Fig. 2B). Table 1 shows the means and the mean differences in the final age estimates (for the census in 2007 and in 2015) according to the three proposed methods which all showed high variations between the tree ages in each population due to the occurrence of old trees (high standard deviation). When the raw results corrected for bark thickness and errors caused by missing/false rings, the tree age was adjusted to –

82 Demography and recruitment of Moringa peregrina Publications

5 years in case of the error caused by bark thickness and to ±3 years in case of missing/ false rings. However, the error due to the different estimations of diameter (in the three proposed methods) may lead to a difference in the estimated age in between 2 to 7 years. Correction for bark thickness decreases age by an average of 3.4±1.2 years, whereas correction for false rings out will increase the age by 3.1±1.8 years. By comparing the similar age cross-cuts, we recorded false rings in 29 trees (22 had only one false ring, six trees had two, and one isolated tree at W. Meir showed three false rings). Differences in mean trees ages between the four wadis were statistically significant (P=0.000), with a high SD for the trees of W. Zaghra. The mean annual increment area of all trees was to be 26.9±17.7 cm2, and the annual increment of tree radius was 1.85±0.12 mm. Correlation between the estimated M. peregrina tree ages and the measured tree size (Table 2) showed a dependent positive relationship between tree age and height (correlation coefficient r = +0.63), diameter at breast height (r = +0.94).

Fig. 3. A Linear regression for the relationship between total radius and bark thickness (left). B Relationship between radius (corrected for bark thickness according to the function derived from A) and the number of annual rings for 117 M. peregrina wood samples.

Table 1. Age estimates (mean±SD) of M. peregrina at the four wadis and the pooled population corrected for the errors caused by bark thickness and false/missing rings.

Final Estimated Age * trees

Year Wadi # [Method A: mean of all radii] [Method B: virtual single radius] [Method C: radius of largest stem] WA 35 55.7±34.2 61.6±35.6 50.9±33.8 WF 45 55.6±32.2 57.6±39.1 46.3±28.5

WM 242 87.6±56.5 102.5±69.2 74.6±45.9

2007 WZ 71 105.6±67.4 122.9±73.4 94.7±63.8 PP 393 84.4±57.0 97.4±67.9 72.9±49.2 WA 35 63.5±34.4 69.8±35.6 56.9±34.5 WF 45 63.4±32.4 65.9±39.1 52.7±28.5

WM 242 95.4±56.7 110.7±69.0 80.2±45.1

2015 WZ 71 113.4±67.9 131.4±73.5 100.3±64.6 PP 393 92.1±57.3 105.7±67.9 78.6±49.0 * Number of trees in 2007 was 406 but all dead and uprooted trees are not included.

Table 2. Correlation between the tree ages of M. peregrina at the four wadis calculated by the methods A: mean of all radii, B: virtual single radius, and C: the radius of largest/oldest stem (in 2007 and 2015) and tree size parameters (height, diameter at breast height and crown cover area) and vitality. H_07 H_15 DBH_07 DBH_15 CrowA_07 CrowA_15 Vit_07 Vit_15 Age_07_A 0.593 0.582 0.894 0.897 0.331 0.335 0.113 0.202 Age_15_A 0.591 0.579 0.892 0.895 0.329 0.333 0.111 0.200 Age_07_B 0.638 0.625 0.940 0.943 0.356 0.359 0.112 0.219 Age_15_B 0.639 0.626 0.940 0.943 0.356 0.359 0.111 0.218 Age_07_C 0.581 0.568 0.867 0.865 0.296 0.301 0.122 0.232 Age_15_C 0.572 0.559 0.857 0.856 0.289 0.294 0.121 0.229

83 Publications Demography and recruitment of Moringa peregrina

3.3. Age structure and size structure Figure 3 shows the demographic status of M. peregrina (the four populations are pooled) based one size (classes of 10 cm dbh) and age structure (classes of 20 years) calculated according to three different methods (A: mean of all radii, B: virtual single radius, and C: the radius of largest/oldest stem). W. Zaghra harbours the oldest trees, followed by W. Meir. For the ages estimated via methods A (mean of all radii) and B (virtual single radius), we did not record any tree in the first age class (younger than 20 years). However, for method C (radius of the largest/oldest stem), one a single tree (19 years) from wadi Zaghra appears in this class. The oldest tree in the studied populations was estimated at 386 years and would have been established at W. Zaghra in 1629. The youngest tree was 27 years (W. Feiran, both trees single-stemmed). Table 3 shows the mean ages of the trees for the censuses of 2007 and 2015, resulting in a mean age difference of 8.25±0.5 years. Figure S4 reveals numerous recruitment gaps within the last 200 years in W. Agala and W. Feiran, and within the last 400 years for W. Meir and Zaghra. Weibull analyses indicated three periods of lacking recruitment within the years 1905– 1925, 1955–1975 and 1990–2015, the last 25 years.

Fig. 3. On the left side, a size and age structures in 2015, expressed as diameter at ground height (DAG, classes of 10 cm diameter, upper left: grey bars) and the estimated age structure (in classes of 20 years, lower left: black bars) for 393 trees of M. peregrina trees at the Sinai Peninsula. The census was based on to the three methods for age estimation in multistemmed trees, the DAG has calculated from the mean of the radii for all stems (A), from a virtual tree radius calculated from the sum of the areas for each stem (B) and from the radius of the largest /oldest stem (C). On the right side, a box plot of the estimated tree ages at the four wadis (Meir WM, Agala WA, Feiran WF, and Zaghra WZ) in Sinai (census 2015).

3.4. Life table of the four Moringa populations Table S1 shows the parameters of the life table for the four populations, including age-specific mortality rates (100% chance of death, qx = 1). According to methods A and B, no trees in W. Agala and W. Feiran are older than 180 years. Trees older than 140 years face the same risk of the dead according to method A at W. 84 Demography and recruitment of Moringa peregrina Publications

Feiran and method C at W. Agala. The oldest trees at W. Meir range from 240 to 320 years and are up to 380

years at W. Zaghra (Table S1). The number of trees (N0-20) established in the last 20 years is zero for all wadis. Because about 30% of the trees at the four wadis did not produce seeds in any of the two censuses, it was difficult to assess the fecundity.

Table 3. Summary statistics of the estimated M. peregrina ages (yrs) with the three methods: A: mean of all radii, B: virtual single radius, and C: the radius of largest/oldest stem (in 2007 and 2015) and the mean age difference.

Estimated ages [yrs] 2007 Estimated ages [yrs] 2015 Age difference [yrs] Method A Method B Method C Method A Method B Method C Method A Method B Method C Mean 84.4 97.4 72.9 92.1 105.7 78.6 8 8 6 St. Deviation 57.03 67.93 49.21 57.26 67.89 48.95 1.02 0.52 1.97 Standard Error 2.9 3.4 2.5 2.9 3.4 2.5 0.05 0.03 0.10 Minimum 18 18 13 27 27 19 6 7 2 Maximum 377 377 377 386 386 386 10 9 9 Median 65.22 76.62 58.15 72.82 83.49 62.86 7.86 8.23 5.50 Mode 48.72 43.22 33.00 58.93 51.86 46.36 7.86 7.86 3.93 Kurtosis 4.1 2.4 5.7 4.1 2.3 5.9 -0.3 -1.1 -1.2 Skewness 1.8 1.5 2.0 1.8 1.5 2.0 -0.5 0.2 0.1

3.5. Survivorship curve and population hazard As for most species of trees, the survivorship curve for Moringa represents type III of survivorship curves: both young and old individuals of M. peregrina at Sinai face higher rates of mortality than middle-aged ones (Fig. 4). Especially the young trees experienced the lowest age-specific survival. This points to the low

establishment and a high mortality rate for saplings. The value of Survivorship lx starts to decline in the four wadis in the age class 21–40 yrs for all three methods of age estimation. However, the decrease in survivorship comes faster at W. Feiran which harbours the youngest trees.

Fig. 4. Survivorship curves (Kaplan-Meier) for M. peregrina calculated by three different methods (A, B, and C) for the four populations (Agala, Feiran, Meir, and Zaghra) and the pooled population in Sinai.

3.6. Mis-Recruitment Index Based on the number of failure years (no tree is added to the population), the mis-recruitment index indicates the proportion of failure years based on the lifespan of the oldest tree. An example is shown in Table S2. The highest percentage of failure years was recorded for Moringa trees of W. Agala with 27.2, 32.7, and 31.6% according to methods A, B, and C, respectively. It is followed by W. Zaghra, with figures of 24.2, 17.5, and 85 Publications Demography and recruitment of Moringa peregrina

32.4%. However, the highest proportions of failure years at W. Feiran were 16.9%, 15.2%, and 16.2 % (method A, B, and C), and at W. Meir was ca. 10% (methods A and B) and 14.6 % (method C). Figure S4 shows the recruitment for the four populations in classes of 10 years. All three methods pointed to longer periods of recruitment failure, with 62 years (1839–1901) and the last 32 years (1983–2015) for W. Agala. Similar long periods are indicated for the other wadis: W. Feiran, 28 years (1841–1869) and the last 27 years (1988–2015); W. Meir 26 years (1725–1750) and the last 36 years (1976–2015), although it comprises the largest population in Sinai (242 trees); W. Zaghra 68 years (1659–1727) and the last 30 years (1985–2015).

4. Discussion According to Vlam et al. (2014), many tropical tree species have population structures that exhibit longer periods of recruitment failure. For the old trees at Sinai area like M. peregrina trees, a trend of rapid climate change (Dadamouny and Schnittler 2016) led to even longer failure periods. Although the wood of M. peregrina is soft and the formation of growth rings could be absent (Olson and Carlquist, 2001) or not clear in the cultivated trees in the botanical gardens (Elwakeel et al. 2017), due to the high aridity of its habitats in Sinai (Dadamouny and Schnittler 2016) and the phenological changes over the year, the trees are producing a growth ring that could be easily counted. This because trees in constantly watered environments do not show ring patterns which could not lead to the conclusion that trees suffer when those are building the rings. Climate

changes; mainly increasing temperature and CO2 level (Van Der Sleen et al. 2015) and the seasonal variations (Yang et al. 2017) and the phenological phases play a role in cambial dormancy and in starting of the growing season and the ring formation (Villalba et al. 1998 and 2011; Brienen et al. 2016; Ohashi et al. 2016; Pérez‐de‐ Lis et al. 2016). To find the optimal relationship between tree age and radius, several measurements were needed to reduce the errors in the estimated age and obtain the trend of recruitment for the remaining M. peregrina populations over time. Since the results showed age difference (7.8±1.0 yrs for method A and 8.2±0.5 yrs for method B) obtained from the differences in the estimated age for the censuses 2007 and 2015 and correlated with the time between the two censuses (8 yrs), we can conclude that M. peregrina produces indeed regular rings (one ring per year). However, the high proportion of multi-stemmed trees poses problems since no diameter at breast height dbh or diameter at ground DAG can be measured above ground level (Powell, 2005). Since M. peregrine grows preferentially at the margin of the wadi beds, where water supply is better but trees still face a significant risk of damage by flash floods (Dadamouny et al. 2016), trees with a broken or uprooted stem are likely to produce multiple stems at the trunk base or from emerging roots. This resulted in a multi-stemmed tree with a solid old trunk protected by the boulders in the wadi bed. Age of these multi-stemmed trees is hard to determine, and from this reason, we applied three different methods (average radius, virtual radius, and radius of the largest stem) in parallel, In addition, the accuracy of the age estimation depends on the proper estimation of bark thickness. Figures for bark thickness vary for the same trunk if measured at different angles. Therefore, we used the mean of measurements at eight angles as an input for fitting the correlation function between trunk diameter and bark diameter. Our estimates depend on further from the number of the collected wood samples, which we limited to dead or broken branches. For the collected ninety-three cross-cuts by Dadamouny (2009), the regression equation was (Age = 5.6822 × radius + 0). For this study, we used 117 wood samples, which slightly changed figures (Age = 5.5851 × [radius ₋ bark] + 0). These deviations would equal an error in age estimation between 2 to 5%. Finally, for an area like the Sinai with extreme fluctuations in annual precipitation 86 Demography and recruitment of Moringa peregrina Publications

(Dadamouny and Schnittler 2016), there was the risk of recording false or missing rings. Due to frequent damage of the trees and climatic stress, M. peregrina trees may produce a false ring (Schulman 1938), similar to tropical trees, where landscape fragmentation and dry hot winds hamper the tree to grow (Venegas-González et al. 2017, Vlam et al., 2017). And similar to the low-density wood trees (e.g. Adansonia), M. peregrina wood would require over-proportionally large corers which will severely harm the tree. We thus had to rely on size - age relationships for estimating the trend of recruitment in the current study. Age structure of the M. peregrina populations was estimated first by Dadamouny (2009) and Zaghloul et al. (2012) based on the mean circumference of the root collar. However, the root collar may be much thicker than a regular trunk (Chojnacky 1999), causing overestimation of age. The three methods applied now are all not free of over- or underestimation, but the outcome is surprisingly similar. By comparing the estimated ages for the censuses 2007 and 2015, we found that the mean difference in the tree ages agrees well with the expectation for growing one ring per year (8.2±0.5 yrs). It confirms that tree size and growth rings could be used as demographic parameters for M. peregrina trees. In addition, sample size plays a critical role in the accuracy of crossdating, used to detect false rings. The initial estimation by Dadamouny (2009) and Zaghloul et al. (2012) resulted in a mean age of the same studied 393 M. peregrina trees of 69.5±47.1 yrs. This is still in agreement with the values of the mean ages for the measured 393 trees based on estimates from 117 cross cuts. However, the mean age differences for method A (mean of all radii) was 7.79±1.0 yrs, B (virtual single radius) was 8.25±0.5 yrs, and C (largest/oldest stem) was 5.71±1.97 yrs. Method A (mean of radii) seems to cause underestimation of age in multistemmed trees since small-sized trunks reduce the average figure for the radius. For such trees, extremely small stems should be excluded. However, method B (virtual single radius by calculating the cross area for each stem) exhibits the actual tree increment and then more precise age estimation. Finally, method B with virtual single radius avoided the underestimation in both methods A (mean of radii) and C (largest stem) and avoided the overestimation of age for the multistemmed trees. Fortunately, by comparing cross sections of the same size, false rings were rarely observed, most often only once (in 22 of 29 trees), but they occurred three times in the oldest wood sample (47 years) collected from W. Meir. By cross-dating estimated the years with false rings to be 1983, 1989 or 1990 and 2009. According to Dadamouny and Schnittler (2016), these years were very dry, where no rainfall recorded at the nearest station (El-Tor). This correlation underpins the importance of fluctuations in annual precipitation on tree performance. The age structure of M. peregrina populations revealed that the majority (54%) of the trees were between 41 and 80 years, and established between 1935 and 1955. Very low recruitment happened from 1955 to 1975. Such fluctuations in recruitment rates are not unusual for desert plants, especially given the fluctuations in rainfall (Dadamouny et al., 2012; Dadamouny and Schnittler, 2016). However, recruitment declined sharply around 20–40 years ago. This may as well be caused by an increase of the anthropogenic impacts through growth in human populations after the end of the 1973 war between Egypt and Israel. The youngest tree established at W. Feiran in 1988, which one of the years with the highest annual precipitation (64 mm, Dadamouny and Schnittler, 2016). After strong rains in 2010, we observed 12 new individuals but these could not be found in the 2015 census. Especially within the last three to four decades recruitment was nearly missing: Around 91% of the trees in W. Agala, 76% in W. Feiran, 98% in W. Meir and 96% in W. Zaghra, and 95% of the pooled population are older than forty years. In other words, all new individuals established during the last forty years make up less than 5% of the entire population. These results are in agreement with statements from other observations (Hegazy et al. 2008, Dadamouny 2009, and Zaghloul et al. 2012): If the current situation remains

87 Publications Demography and recruitment of Moringa peregrina

unchanged, the populations of M. peregrina at the Sinai face a high risk of extinction. As the older trees are not being replaced by young trees. Another problem is low survival as revealed by a life table and Kaplan-Meier analysis. M. peregrina trees at W. Agala and Feiran have a higher mortality rate than those of the other two wadis (Meir and Zaghra). To maintain the populations we need to identify sites were the remaining trees perform well, and facilitated recruitment may have a chance. Since we obtained a type III survivorship curve, characterized by a high rate of mortality among both young and old trees, the population will not remain stable if the current period of low recruitment lasts. 5. Recommendations The regeneration and conservation programs (preferably overseen by local Bedouins) should focus on facilitating recruitment of young individuals in selected natural habitats. This program should include as well the protection of the old trees to preserve the genetic diversity within the trees of the area. Collecting and germinating seeds from the largest and oldest trees in W. Zaghra and W. Meir, including the oldest one (located at R36 626386 E and 3170610 N according to UTM), maybe the first step. Because we based our estimation of the trees age on two censuses only (2007 and 2015), a long-term monitoring of population structure would be required. Together with records on years with extreme weather events (drought or local flash floods), this would enable a population viability analysis (PVA) which is urgently needed to identify the drivers of population dynamics for the tree. We recommend as well to update the threat assessment for M. peregrina in the IUCN Red List of Threatened Species. Since M. peregrina is threatened to several factors, mainly anthropogenic impact and climate change, which may act in a synergistic manner, it is hard to predict the risk of extinction. To minimize risks, an action plan by the Saint Catherine Protectorate Authority would help in mitigating the adverse impacts of these threats. A hitherto unknown factor is the genetic diversity of the populations and its influence on survival. We thus recommend further studies on reproductive biology and genetic diversity within and among these populations. 6. Conclusions M. peregrina is one of the few subtropical trees displaying annual growth rings that allow age estimation. Due seasonal changes in Sinai as a hyperarid area, and due to the pronounced phenological changes of M. peregrina, and correlation between the tree size and age in the two different seasons, the tree produces a growth ring per year. This is supported also by the estimated mean age difference from two censuses (2007 and 2015), employing the relationship between the radius corrected for bark thickness and annual growth rings derived from 117 collected cross-cuts. The three different methods used for age estimation produced comparable results and revealed longer periods of recruitment failure for all locations of the tree. In combination with human overuse and climate change especially in the last four decades, natural fluctuations in recruitment pose a high risk for the remaining populations that are in severe decline. For these reasons, we recommend an action plan and updating the threat assessment of M. peregrina on the IUCN Red List from Vulnerable (VU) to Endangered (EN). Acknowledgments The first author appreciates the support by the German Academic Exchange Service (DAAD) in the frame of the German-Egyptian Research Long-term Scholarship program (GERLS). Our appreciation extends to the Ministry of Higher Education, Egypt and the authorities of St. Catherine Protectorate for their effort in protecting the endangered species in the area. 88 Demography and recruitment of Moringa peregrina Publications

Supplementary data: Appendices for the supplementary data associated with this article are available online at: http://dx.doi.org/10.17632/z23sx6vjc3.1 References Batcheler, C.L., 1985. Note on measurement of woody plant diameter distributions. J. Ecol. 8,129–132, 1985. Bengon, M. Harper, J.L. and Townsend, C.R., 1996. Ecology: Individuals, Populations, and Communities, 3rd Ed. Blackwell Science Ltd. Cambridge, Mass. 1068 pp. https://doi.org/10.1002/9781444313765. Boukili, V.K., Bebber, D.P., Mortimer, T., Venicx, G., Lefcourt, D., Chandler, M., Eisenberg, C. 2017. Assessing the performance of urban forest carbon sequestration models using direct measurements of tree growth. Urban Forestry and Urban Greening, 24, 212–221. https://doi.org/10.1016/j.ufug.2017.03.015. Bouriaud, O., Leban, J.M., Bert, D., and Deleuze, C., 2005. Intra-annual variations in climate influence growth and wood density of Norway spruce. Tree Physiol, 25, 651–660. https://doi.org/10.1093/treephys/25.6.651 Brienen R.J., Schöngart, J., Zuidema, P.A. 2016. Tree rings in the tropics: insights into the ecology and climate sensitivity of tropical trees. In Tropical tree physiology (pp. 439–461). Springer, Cham. https://doi.org/10.1007/978-3-319- 27422-5_20. Chojnacky, D.C., 1999. Converting Tree Diameter Measured at Root Collar to Diameter at Breast Height, WJAF (1), 14– 16. Copenheaver, C.A., Pokorski, E.A., Currie, J.E. and Abrams, M.D., 2006. Causation of false ring formation in Pinus banksiana: A comparison of age, canopy class, climate and growth rate.Forest Ecol. and Management, 236, 348– 355. https://doi.org/10.1016/j.foreco.2006.09.020. Coster, C. 1927. Zur Anatomie und Physiologie der Zuwachszonen-und Jahresringbildung in den Tropen, Annales Jardim Botanica Buitenzorg, 37, 49–161. Cox, D.R., 1972. Regression Models and life tables. J. of the Royal Statistical Society, Series B, 34,187–220. Dadamouny, M.A. 2009. Population Ecology of Moringa peregrina growing in Southern Sinai, Egypt. M.Sc. Thesis, Environmental Sciences, Faculty of Science, Suez Canal University, Egypt, 205 pp. Dadamouny, M.A., Schnittler, M. 2016. Trends of climate with rapid change in Sinai, Egypt. J water and climate change 7(2), 393–414 https://doi.org/10.2166/wcc.2015.215. Dadamouny M.A., Unterseher M., König, P., Schnittler, M. 2016. Population performance of Moringa peregrina (Forssk.) Fiori (Moringaceae) at Sinai Peninsula, Egypt in the last decades: consequences for its conservation. Journal for Nature Conservation 34, 65–74 http://dx.doi.org/10.1016/j.jnc.2016.08.005. Dadamouny, M.A., Zaghloul, M.S., Moustafa, A.A., 2012. Impact of Improved Soil Properties on Establishment of Moringa peregrina Seedlings and a Trial to Decrease its Mortality Rate. Egy J Bot 52(1), 83–98 de Avila A.L., Schwartz G., Ruschel, A.R., do Carmo L.J., Silva, J.M., de Carvalho, J.P., Dormann C.F., Mazzei L., Soares M., Bauhus J. (2017). Recruitment, growth and recovery of commercial tree species over 30years following logging and thinning in a tropical rainforest. Forest Ecology and Management 385, 225–235. https://doi.org/10.1016/j.foreco.2016.11.039. Dean, C., 2003. Calculation of wood volume and stem taper using terrestrial single-image close-range photogrammetry and contemporary software tools. Silva Fennica 37(3), 359–380. https://doi.org/10.14214/sf.495. Elwakeel, A.O., Osman, H.E., Abohassan, A.A. 2017. Anatomical Characteristics of the Stem Wood of Moringa peregrina (Forssk.) Fiori Planted in Western Saudi Arabia. International Journal of Theoretical and Applied Sciences, 9(2), 205–211. Ferguson, C.W. 1970. Concepts and techniques of dendrochronology. In Berger, R. editor, Scientific methods in medieval archeology, Berkeley: University of California. American J Science 233, 140–46 Fritts, H.C., Swetnam, T.W. 1989. Dendroecology: a tool for evaluating variations in past and present forest environments. In Advances in ecological research 19, 111–188. Academic Press. https://doi.org/10.1016/S0065-2504(08)60158-0 Fritts, H.C., Vaganov, E.A., Sviderskaya, I.V., Shashkin, A.V. 1991. Climatic variation and tree-ring structure in conifers: empirical and mechanistic models of tree-ring width, number of cells, cell size, cell-wall thickness and wood density. Climate Research, 97–116. https://doi.org/10.3354/cr001097. Harper, J.L., White, J. 1974. The demography of Plants. Annual Review of Ecology and Systematics 5, 419–463. https://doi.org/10.1146/annurev.es.05.110174.002223 Hegazy, A.K., Hammouda, O., Lovett-doust, J., Gomaa, N.H. (2008). Population dynamics of Moringa peregrina along an altitudinal gradient in the northwestern sector of the Red Sea. J Arid Environments 72, 1537–155. https://doi.org/10.1016/j.jaridenv.2008.03.001. 89 Publications Demography and recruitment of Moringa peregrina

Jacquin P., Longuetaud F., Leban J.M., Mothe F 2017. X-ray microdensitometry of wood: a review of existing principles and devices. Dendrochronologia, 42, 42–50. https://doi.org/10.1016/j.dendro.2017.01.004. Lawerence, D. B., 1950. Estimating dates of recent glaciers advances and recession rates by studying tree growth layers. Transaction of the American Geophysical Union 31, 243–48. https://doi.org/10.1029/TR031i002p00243. Lexerød, N., Eid, T. 2005. Recruitment models for Norway spruce, Scots pine, birch and other broadleaves in young growth forests in Norway. Silva Fennica 39(3), 391–406. https://doi.org/10.14214/sf.376. Lieberman, D., Lieberman, M., Hartshorn, G., Peralta, R. 1985. Growth rates and age-size relationships of tropical wet forest trees in Costa Rica. Journal of tropical Ecology, 1(2), 97–109. https://doi.org/10.1017/S026646740000016X. Molles J, Manual C (2002). Ecology, Concepts and Applications. 2nd ed., McGraw-Hill Higher Education, New York, 586 pp. NCSS (Statistical System) 2015. Statistical, graphics, and sample size software NCSS 10 http://www.ncss.com/ Ogden, J., 1981. Dendrochronological studies and determination of tree ages in the Australian tropics. J Biogeography 8, 405–420. https://doi.org/10.2307/2844759. Ohashi, S., Durgante, F.M., Kagawa, A., Kajimoto, T., Trumbore, S.E., Xu, X., Higuchi, N. 2016. Seasonal variations in the stable oxygen isotope ratio of wood cellulose reveal annual rings of trees in a Central Amazon terra firme forest. Oecologia, 180(3), 685–696. https://doi.org/10.1007/s00442-015-3509-x. Olson, M.E. 2002. Combining data from DNA sequences and morphology for a phylogeny of Moringaceae. Syst. Bot. 27, 55–73. Olson, M.E., Carlquist, S. 2001. Stem and root anatomical correlations with life form diversity, ecology, and systematics in Moringa (Moringaceae). Botanical Journal of the Linnean Society, 135(4), 315–348. https://doi.org/10.1111/j.1095-8339.2001.tb00786.x. Pérez‐de‐Lis G., Rossi S., Vázquez‐Ruiz R.A., Rozas V., García‐González, I. 2016. Do changes in spring phenology affect earlywood vessels? Perspective from the xylogenesis monitoring of two sympatric ring‐porous oaks. New Phytologist, 209(2), 521–530. https://doi.org/10.1111/nph.13610. Powell, D.C. 2005. How to measure a big tree. Umatilla National Forest, Online Notes, 1-9. Available: http://www.fs.usda.gov/Internet/FSE_DOCUMENTS/stelprdb5202838.pdf Roque, R.M., Tomazelo, Filho, M. 2015. Relationships between anatomical features and intra-ring wood density profiles in Gmelina arborea applying X-ray densitometry. Cerne, 13(4), 384–392. Rozendaal, D.M., Zuidema, P.A. 2011. Dendroecology in the tropics: a review. Trees, 25(1), 3–16. https://doi.org/10.1007/s00468-010-0480-3. Santos, G.M., Linares R., Lisi C.S., Tomazello Filho, M. 2015. Annual growth rings in a sample of Paraná pine (Araucaria angustifolia): toward improving the 14C calibration curve for the Southern Hemisphere. Quaternary Geochronology, 25, 96–103. https://doi.org/10.1016/j.quageo.2014.10.004. Schulman, E. 1938. Classification of false annual rings in Monterey pine. Tree-Ring Bulletin 4(3), 4–7. Schweingruber, F.H., 1988. Tree rings: Basics and Applications of Dendrochronology. Kluwer Academic Publishers, ISBN 9027724458, 276 pp. https://doi.org/10.1007/978-94-009-1273-1. Schweingruber, F.H. 2012. Trees and wood in dendrochronology: morphological, anatomical, and tree-ring analytical characteristics of trees frequently used in dendrochronology. Springer Science & Business Media, 401 pp Sharitz, R.B., McCormick, J.F., 1973. Population dynamics of two competing annual plant species. Ecology 54, 723–740. https://doi.org/10.2307/1935669. Silvertown, J.W. 1982. Introduction to plant population ecology. Longman Group Limited. London, New York, 209 pp Soltan, M.M., Zaki, A.K. 2009. Antiviral screening of forty-two Egyptian medicinal plants. J ethnopharmacology 126(1), 102–107. https://doi.org/10.1016/j.jep.2009.08.001. Sperry, J.S., Saliendra, N.Z. 1994. Intra‐and inter‐plant variation in xylem cavitation in Betula occidentalis. Plant, Cell & Environment, 17(11), 1233–1241. https://doi.org/10.1111/j.1365-3040.1994.tb02021.x. Stovall, A.E., Vorster, A.G., Anderson, R.S., Evangelista, P.H., Shugart, H.H. 2017. Non-destructive aboveground biomass estimation of coniferous trees using terrestrial LiDAR. Remote Sensing Env, 200, 31–42. https://doi.org/10.1016/j.rse.2017.08.013. Van Der Sleen, P., Groenendijk, P., Vlam, M., Anten, N.P., Boom, A., Bongers, F., Zuidema, P.A. 2015. No growth

90 Demography and recruitment of Moringa peregrina Publications

stimulation of tropical trees by 150 years of CO2 fertilization but water-use efficiency increased. Nature geoscience, 8(1), 24–28 https://doi.org/10.1038/ngeo2313. Venegas-González, A., Brancalion, P.H., Albiero, A., Chagas, M.P., Anholetto, C.R., Chaix, G., Tomazello Filho, M. 2017. What tree rings can tell us about the competition between trees and lianas? A case study based on growth, anatomy, density, and carbon accumulation. Dendrochronologia 42, 1–11. https://doi.org/10.1016/j.dendro.2016.11.001. Villalba, R., Grau, H.R., Boninsegna, J.A., Jacoby, G.C., Ripalta, A. 1998. Tree‐ring evidence for long‐term precipitation changes in subtropical South America. International Journal of Climatology, 18(13), 1463–1478. https://doi.org/10.1002/(SICI)1097-0088(19981115)18:13<1463::AID-JOC324>3.0.CO;2-A Villalba, R., Luckman, B.H., Boninsegna J., D’Arrigo R.D., Lara, A., Villanueva-Diaz J., Pastur G.M. 2011. Dendroclimatology from regional to continental scales: Understanding regional processes to reconstruct large-scale climatic variations across the western Americas. In MK Hughes, TW Swetnam, HF Diaz (Eds.), Dendroclimatology: Progress and prospects (pp. 175–227). New York, NY: Springer. https://doi.org/10.1007/978-1-4020-5725-0_7. Vlam, M., van der Sleen, P., Groenendijk, P., Zuidema, P.A. 2017. Tree Age Distributions Reveal Large-Scale Disturbance-Recovery Cycles in Three Tropical Forests. Article 1984, Frontiers in Plant Science, 7, 1–12. https://doi.org/10.3389/fpls.2016.01984. Worbes, M. 1999. Annual growth rings, rainfall‐dependent growth and long‐term growth patterns of tropical trees from the Caparo Forest Reserve in Venezuela. Journal of ecology, 87(3), 391–403. https://doi.org/10.1046/j.1365- 2745.1999.00361.x. Worbes, M., Staschel, R., Roloff, A., Junk, W.J. 2003. Tree ring analysis reveals age structure, dynamics and wood production of natural forest stand in Cameron. J Forest Ecology and Management, 173(1-3), 105–123. https://doi.org/10.1016/S0378-1127(01)00814-3. Yang, B., He, M., Shishov, V., Tychkov, I., Vaganov, E., Rossi, S., Grießinger, J. 2017. New perspective on spring vegetation phenology and global climate change based on Tibetan Plateau tree-ring data. Proceedings of the National Academy of Sciences, 114(27), 6966–6971. https://doi.org/10.1073/pnas.1616608114. Zaghloul, M.S., Moustafa, A.A., Dadamouny, M.A. 2012. Age structure and population dynamics of Moringa peregrina, an economically valuable medicinal plant. Egyptian J Botany 52(2), 499–519. Zahran, M.A., Willis, A.J. 2009. The vegetation of Egypt. 2nd Ed., London: Chapman and Hall 437 pp. Zoltai, S.C., 1975. Tree-ring record of soil movements on permafrost. Arctic and Alpine Research 7, 331–34. https://doi.org/10.2307/1550177.

91 Appendix 05 Appendix V

Demography and recruitment in populations of the threatened tree Moringa peregrina (Moringaceae) at Sinai Peninsula, Egypt

Mohamed A. Dadamouny and Martin Schnittler University Greifswald, Institute of Botany∗ and Landscape Ecology, Soldmannstr. 15, D-17487 Greifswald, Germany * Corresponding author. E-mail: [email protected], [email protected]. Fax: +49 3834 86 41 14

Fig. S1. Moringa tree in W. Meir showing multiple stems (a), measuring its CAG (b), and coring trial for one tree in W. Feiran (c).

Fig. S2. Collected Moringa cross cuts (a) showing patterns of growth rings and the bark thickness (b) and the radius of wood without the bark thickness (c) as a result of eight diameters divided by two. 92

Appendix 05

Fig. S3. A Scheme of a multistemmed tree showing the problem of measuring the tree diameter DAG, since the real trunk is hidden in the ground. B Numbers and percentage of multistemmed trees in the four wadis (Agala WA, Feiran WF, Meir WM, Zaghara WZ).

Fig. S4. Recruitment of Moringa trees as the number of newly arising trees according to age estimates in the four wadis (classes of 10 years), derived from age estimations based on three methods (blue for method A: average all radii; green for method B: virtual single radius calculated total cross area; and orange for method C: largest/oldest stem).

Available online: http://dx.doi.org/10.17632/z23sx6vjc3.1 93

Appendix 06 Appendix VI

Table S1. Age-specific mortality rates qx for static life tables calculated for four Moringa peregrina populations at Sinai. Variables are x = age entered by the time of the census; Nx = a number of individuals living in age class x; ax = a number of individuals that survive to age x; survivorship lx = proportion of original cohort surviving to age x; Lx = the average proportion of trees being alive at that age; Tx = the total number of living individuals at age class x and beyond; ex = the probability of a tree living 'x' years beyond a given age; dx = number of individuals that die during the time interval corresponding to age class x; qx = proportion of individuals entering age x that die during that period. The table is calculated based on the estimated ages with the three methods A, B, and C. Method A: mean of total radii of multistemmed trees Age Wadi Agala WA Wadi Feiran WF Wadi Meir WM Wadi Zaghra WZ

Class NRx aRx lRx LRx TRx eRx dRx qRx NRx aRx lRx LRx TRx eRx dRx qRx NRx aRx lRx LRx TRx eRx dRx qRx NRx aRx lRx LRx TRx eRx dRx qRx [≤ 20] 0 35 1.0 1.0 3.2 3.2 0.0 0.0 0.0 45.0 1.0 1.0 3.1 3.1 0.0 0.0 0 242 1.0 1.0 4.8 4.8 0.0 0.0 0.0 71.0 1.0 1.0 5.6 5.6 0.0 0.0 [21-40] 5 35 1.0 0.9 2.2 2.2 5.0 0.1 12.0 45.0 1.0 0.9 2.1 2.1 12.0 0.3 11 242 1.0 1.0 3.8 3.8 11.0 0.0 4.0 71.0 1.0 1.0 4.6 4.6 4.0 0.1 [41-60] 16 30 0.9 0.6 1.3 1.5 16.0 0.5 16.0 33.0 0.7 0.6 1.3 1.7 16.0 0.5 76 231 1.0 0.8 2.8 2.9 76.0 0.3 12.0 67.0 0.9 0.9 3.7 3.9 12.0 0.2 [61-80] 8 14 0.4 0.3 0.6 1.6 8.0 0.6 5.0 17.0 0.4 0.3 0.7 1.9 5.0 0.3 36 155 0.6 0.6 2.0 3.1 36.0 0.2 9.0 55.0 0.8 0.7 2.8 3.6 9.0 0.2 [81-100] 3 6 0.2 0.1 0.3 2.0 3.0 0.5 4.0 12.0 0.3 0.2 0.4 1.5 4.0 0.3 35 119 0.5 0.4 1.4 2.9 35.0 0.3 10.0 46.0 0.6 0.6 2.1 3.2 10.0 0.2 [101-120] 1 3 0.1 0.1 0.2 2.5 1.0 0.3 5.0 8.0 0.2 0.1 0.2 1.0 5.0 0.6 22 84 0.3 0.3 1.0 2.9 22.0 0.3 12.0 36.0 0.5 0.4 1.5 3.0 12.0 0.3 [121-140] 0 2 0.1 0.1 0.1 2.5 0.0 0.0 2.0 3.0 0.1 0.0 0.1 0.8 2.0 0.7 22 62 0.3 0.2 0.7 2.7 22.0 0.4 7.0 24.0 0.3 0.3 1.1 3.3 7.0 0.3 [141-160] 0 2 0.1 0.1 0.1 0.0 0.0 0.0 1.0 1.0 0.0 0.0 0.0 0.5 1.0 1.0 7 40 0.2 0.2 0.5 2.9 7.0 0.2 4.0 17.0 0.2 0.2 0.8 3.4 4.0 0.2 [161-180] 1 2 0.1 0.0 0.0 0.0 1.0 0.5 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 13 33 0.1 0.1 0.3 2.4 13.0 0.4 5.0 13.0 0.2 0.1 0.6 3.3 5.0 0.4 [181-200] 1 1 0.0 0.0 0.0 0.0 1.0 1.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 7 20 0.1 0.1 0.2 2.7 7.0 0.4 2.0 8.0 0.1 0.1 0.5 4.0 2.0 0.3 [201-220] 0 0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 3 13 0.1 0.0 0.2 2.9 3.0 0.2 1.0 6.0 0.1 0.1 0.4 4.2 1.0 0.2 [221-240] 0 0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 3 10 0.0 0.0 0.1 2.6 3.0 0.3 1.0 5.0 0.1 0.1 0.3 3.9 1.0 0.2 [241-260] 0 0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 1 7 0.0 0.0 0.1 2.5 1.0 0.1 1.0 4.0 0.1 0.0 0.2 3.8 1.0 0.3 [261-280] 0 0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 1 6 0.0 0.0 0.0 1.8 1.0 0.2 1.0 3.0 0.0 0.0 0.2 3.8 1.0 0.3 [281-300] 0 0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 3 5 0.0 0.0 0.0 1.1 3.0 0.6 0.0 2.0 0.0 0.0 0.1 4.5 0.0 0.0 [300-320] 0 0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 1 2 0.0 0.0 0.0 1.0 1.0 0.5 0.0 2.0 0.0 0.0 0.1 3.5 0.0 0.0 [321-340] 0 0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 1 1 0.0 0.0 0.0 0.5 1.0 1.0 0.0 2.0 0.0 0.0 0.1 2.5 0.0 0.0 [341-360] 0 0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0 0 0.0 0.0 0.0 0.0 0.0 0.0 1.0 2.0 0.0 0.0 0.0 1.5 1.0 0.5 [361-380] 0 0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0 0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 1.0 0.0 0.0 0.0 1.5 0.0 0.0 [>380] 0 0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0 0 0.0 0.0 0.0 0.0 0.0 0.0 1.0 1.0 0.0 0.0 0.0 0.5 1.0 1.0 Total 35 45 242 71 Method B: virtual single radius Age Wadi Agala WA Wadi Feiran WF Wadi Meir WM Wadi Zaghra WZ

Class NRx aRx lRx LRx TRx eRx dRx qRx NRx aRx lRx LRx TRx eRx dRx qRx NRx aRx lRx LRx TRx eRx dRx qRx NRx aRx lRx LRx TRx eRx dRx qRx [≤ 20] 0 35 1.0 1.0 3.5 3.5 0.0 0.0 0.0 45.0 1.0 1.0 3.3 3.3 0.0 0.0 0.0 242.0 1.0 1.0 5.5 5.5 0.0 0.0 0.0 71.0 1.0 1.0 6.5 6.5 0.0 0.0 [21-40] 3 35 1.0 1.0 2.5 2.5 3.0 0.1 11.0 45.0 1.0 0.9 2.3 2.3 11.0 0.2 3.0 242.0 1.0 1.0 4.5 4.5 3.0 0.0 3.0 71.0 1.0 1.0 5.5 5.5 3.0 0.0 [41-60] 15 32 0.9 0.7 1.5 1.7 15.0 0.5 18.0 34.0 0.8 0.6 1.4 1.9 18.0 0.5 70.0 239.0 1.0 0.8 3.6 3.6 70.0 0.3 8.0 68.0 1.0 0.9 4.6 4.8 8.0 0.1 [61-80] 7 17 0.5 0.4 0.8 1.7 7.0 0.4 6.0 16.0 0.4 0.3 0.8 2.4 6.0 0.4 33.0 169.0 0.7 0.6 2.7 3.9 33.0 0.2 7.0 60.0 0.8 0.8 3.7 4.3 7.0 0.1 [81-100] 6 10 0.3 0.2 0.4 1.5 6.0 0.6 1.0 10.0 0.2 0.2 0.6 2.5 1.0 0.1 29.0 136.0 0.6 0.5 2.1 3.7 29.0 0.2 12.0 53.0 0.7 0.7 2.9 3.8 12.0 0.2 [101-120] 2 4 0.1 0.1 0.2 2.0 2.0 0.5 4.0 9.0 0.2 0.2 0.3 1.7 4.0 0.4 26.0 107.0 0.4 0.4 1.6 3.6 26.0 0.2 7.0 41.0 0.6 0.5 2.2 3.8 7.0 0.2 [121-140] 0 2 0.1 0.1 0.1 2.5 0.0 0.0 2.0 5.0 0.1 0.1 0.2 1.7 2.0 0.4 18.0 81.0 0.3 0.3 1.2 3.5 18.0 0.2 8.0 34.0 0.5 0.4 1.7 3.5 8.0 0.2 [141-160] 0 2 0.1 0.1 0.1 0.0 0.0 0.0 1.0 3.0 0.1 0.1 0.1 1.5 1.0 0.3 12.0 63.0 0.3 0.2 0.9 3.4 12.0 0.2 7.0 26.0 0.4 0.3 1.3 3.4 7.0 0.3 [161-180] 1 2 0.1 0.0 0.0 0.0 1.0 0.5 1.0 2.0 0.0 0.0 0.0 1.0 1.0 0.5 8.0 51.0 0.2 0.2 0.7 3.1 8.0 0.2 5.0 19.0 0.3 0.2 0.9 3.5 5.0 0.3 [181-200] 1 1 0.0 0.0 0.0 0.0 1.0 1.0 1.0 1.0 0.0 0.0 0.0 0.0 1.0 1.0 17.0 43.0 0.2 0.1 0.5 2.6 17.0 0.4 4.0 14.0 0.2 0.2 0.7 3.6 4.0 0.3 [201-220] 0 0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 7.0 26.0 0.1 0.1 0.3 3.0 7.0 0.3 1.0 10.0 0.1 0.1 0.5 3.8 1.0 0.1 [221-240] 0 0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 6.0 19.0 0.1 0.1 0.2 2.9 6.0 0.3 2.0 9.0 0.1 0.1 0.4 3.2 2.0 0.2 [241-260] 0 0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 4.0 13.0 0.1 0.0 0.2 3.0 4.0 0.3 1.0 7.0 0.1 0.1 0.3 2.9 1.0 0.1 [261-280] 0 0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 1.0 9.0 0.0 0.0 0.1 3.1 1.0 0.1 3.0 6.0 0.1 0.1 0.2 2.3 3.0 0.5 [281-300] 0 0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 1.0 8.0 0.0 0.0 0.1 2.4 1.0 0.1 1.0 3.0 0.0 0.0 0.1 3.2 1.0 0.3 [300-320] 0 0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 3.0 7.0 0.0 0.0 0.0 1.6 3.0 0.4 0.0 2.0 0.0 0.0 0.1 3.5 0.0 0.0 [321-340] 0 0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 1.0 4.0 0.0 0.0 0.0 1.5 1.0 0.3 0.0 2.0 0.0 0.0 0.1 2.5 0.0 0.0 [341-360] 0 0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 2.0 3.0 0.0 0.0 0.0 0.8 2.0 0.7 1.0 2.0 0.0 0.0 0.0 1.5 1.0 0.5 [361-380] 0 0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 1.0 1.0 0.0 0.0 0.0 0.5 1.0 1.0 0.0 1.0 0.0 0.0 0.0 1.5 0.0 0.0 [>380] 0 0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 1.0 1.0 0.0 0.0 0.0 0.5 1.0 1.0 Total 35 45 242 71 Method C: Largest /oldest stem Age Wadi Agala WA Wadi Feiran WF Wadi Meir WM Wadi Zaghra WZ

Class NRx aRx lRx LRx TRx eRx dRx qRx NRx aRx lRx LRx TRx eRx dRx qRx NRx aRx lRx LRx TRx eRx dRx qRx NRx aRx lRx LRx TRx eRx dRx qRx [≤ 20] 0 35 1.0 1.0 2.8 2.8 0.0 0.0 0.0 45.0 1.0 1.0 2.7 2.7 0.0 0.0 0.0 242.0 1.0 1.0 4.0 4.0 0.0 0.0 1.0 71.0 1.0 1.0 5.0 5.0 1.0 0.0 [21-40] 14 35 1.0 0.8 1.8 1.8 14.0 0.4 19.0 45.0 1.0 0.8 1.7 1.7 19.0 0.4 19.0 242.0 1.0 1.0 3.0 3.0 19.0 0.1 6.0 70.0 1.0 0.9 4.0 4.1 6.0 0.1 [41-60] 11 21 0.6 0.4 1.0 1.6 11.0 0.5 12.0 26.0 0.6 0.4 0.9 1.6 12.0 0.5 86.0 223.0 0.9 0.7 2.0 2.2 86.0 0.4 12.0 64.0 0.9 0.8 3.1 3.4 12.0 0.2 [61-80] 3 10 0.3 0.2 0.5 1.9 3.0 0.3 6.0 14.0 0.3 0.2 0.5 1.5 6.0 0.4 53.0 137.0 0.6 0.5 1.3 2.3 53.0 0.4 12.0 52.0 0.7 0.6 2.3 3.1 12.0 0.2 [81-100] 4 7 0.2 0.1 0.3 1.5 4.0 0.6 4.0 8.0 0.2 0.1 0.2 1.3 4.0 0.5 30.0 84.0 0.3 0.3 0.8 2.4 30.0 0.4 14.0 40.0 0.6 0.5 1.6 2.9 14.0 0.4 [101-120] 1 3 0.1 0.1 0.2 1.8 1.0 0.3 2.0 4.0 0.1 0.1 0.1 1.0 2.0 0.5 21.0 54.0 0.2 0.2 0.5 2.4 21.0 0.4 9.0 26.0 0.4 0.3 1.2 3.2 9.0 0.3 [121-140] 1 2 0.1 0.0 0.1 1.5 1.0 0.5 2.0 2.0 0.0 0.0 0.0 0.0 2.0 1.0 9.0 33.0 0.1 0.1 0.4 2.6 9.0 0.3 3.0 17.0 0.2 0.2 0.9 3.6 3.0 0.2 [141-160] 0 1 0.0 0.0 0.0 0.0 1.0 1.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 6.0 24.0 0.1 0.1 0.2 2.4 6.0 0.3 2.0 14.0 0.2 0.2 0.6 3.2 2.0 0.1 [161-180] 0 1 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 8.0 18.0 0.1 0.1 0.1 2.0 8.0 0.4 4.0 12.0 0.2 0.1 0.5 2.8 4.0 0.3 [181-200] 1 1 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 4.0 10.0 0.0 0.0 0.1 2.2 4.0 0.4 4.0 8.0 0.1 0.1 0.3 2.9 4.0 0.5 [201-220] 0 0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 6.0 0.0 0.0 0.1 2.3 0.0 0.0 1.0 4.0 0.1 0.0 0.2 4.3 1.0 0.3 [221-240] 0 0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 1.0 6.0 0.0 0.0 0.0 1.3 1.0 0.2 0.0 3.0 0.0 0.0 0.2 4.5 0.0 0.0 [241-260] 0 0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 5.0 5.0 0.0 0.0 0.0 0.0 5.0 1.0 0.0 3.0 0.0 0.0 0.1 3.5 0.0 0.0 [261-280] 0 0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 2.0 3.0 0.0 0.0 0.1 2.5 2.0 0.7 [281-300] 0 0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 1.0 0.0 0.0 0.1 5.5 0.0 0.0 [300-320] 0 0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 1.0 0.0 0.0 0.1 4.5 0.0 0.0 [321-340] 0 0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 1.0 0.0 0.0 0.0 3.5 0.0 0.0 [341-360] 0 0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 1.0 0.0 0.0 0.0 2.5 0.0 0.0 [361-380] 0 0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 1.0 0.0 0.0 0.0 0.0 0.0 0.0 [>380] 0 0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 1.0 1.0 0.0 0.0 0.0 0.0 1.0 1.0 Total 35 45 242 71

94

Appendix 06

Table S1. Mis-Recruitment Index and the percentage of failure years based on the estimated M. peregrina ages (yrs) at W. Agala* according to the three methods. A: mean of all radii, B: virtual single radius, and C: the radius of largest/oldest stem. Method A Method B Method C WA_Age_15 Nd Nd/PA RF%_WA Est. Year WA_Age_15 Nd Nd/PA RF%_WA Est. Year WA_Age_15 Nd Nd/PA RF%_WA Est. Year 31 0 31 0.16 16.21 2015 32 0 32 0.17 17.05 2015 24 0 24 0.125 12.47 2015 32 31 1 0.01 0.83 1984 37 32 5 0.03 2.67 1983 27 24 3 0.017 1.66 1991 37 32 5 0.02 2.29 1983 39 37 2 0.01 0.66 1978 27 27 0 0.000 0.00 1988 37 37 0 0.00 0.00 1978 40 39 2 0.01 0.83 1976 27 27 0 0.000 0.00 1988 39 37 2 0.01 1.04 1978 42 40 2 0.01 1.25 1975 31 27 4 0.025 2.49 1988 40 39 2 0.01 0.83 1976 44 42 2 0.01 0.83 1973 32 31 1 0.004 0.42 1984 42 40 2 0.01 1.25 1975 45 44 1 0.00 0.27 1971 35 32 2 0.012 1.25 1983 42 42 0 0.00 0.00 1973 45 45 0 0.00 0.15 1970 36 35 1 0.004 0.42 1980 44 42 2 0.01 0.83 1973 45 45 0 0.00 0.13 1970 37 35 2 0.008 0.83 1980 44 44 0 0.00 0.00 1971 46 45 1 0.00 0.42 1970 38 37 1 0.004 0.42 1978 44 44 0 0.00 0.00 1971 46 46 0 0.00 0.07 1969 38 38 0 0.000 0.00 1977 45 44 1 0.00 0.42 1971 47 46 1 0.01 0.71 1969 39 38 1 0.004 0.42 1977 46 45 1 0.00 0.42 1970 48 47 1 0.00 0.36 1968 39 39 1 0.004 0.42 1976 46 46 0 0.00 0.00 1969 52 48 4 0.02 1.96 1967 39 39 0 0.000 0.00 1976 47 46 2 0.01 0.83 1969 55 52 3 0.01 1.49 1963 40 39 1 0.004 0.42 1976 48 47 1 0.01 0.62 1968 56 55 1 0.01 0.73 1960 42 40 2 0.008 0.83 1975 52 48 4 0.02 2.08 1967 56 56 0 0.00 0.14 1959 42 42 1 0.004 0.42 1973 57 52 5 0.03 2.70 1963 59 56 3 0.01 1.46 1959 44 42 2 0.008 0.83 1973 58 57 1 0.00 0.42 1958 63 59 4 0.02 2.08 1956 45 44 1 0.004 0.42 1971 59 58 1 0.00 0.42 1957 63 63 0 0.00 0.25 1952 46 45 2 0.008 0.83 1970 59 59 0 0.00 0.00 1956 64 63 1 0.00 0.41 1952 47 46 1 0.004 0.42 1969 61 59 2 0.01 1.25 1956 65 64 1 0.01 0.58 1951 54 47 7 0.037 3.74 1968 62 61 0 0.00 0.21 1954 70 65 5 0.02 2.39 1950 54 54 0 0.000 0.00 1961 63 62 1 0.01 0.62 1953 73 70 3 0.02 1.70 1945 58 54 4 0.021 2.08 1961 63 63 0 0.00 0.00 1952 74 73 1 0.01 0.81 1942 59 58 1 0.004 0.42 1957 64 63 1 0.01 0.70 1952 84 74 9 0.05 4.94 1941 63 59 4 0.021 2.08 1956 65 64 1 0.01 0.55 1951 85 84 1 0.01 0.57 1931 65 63 2 0.012 1.25 1952 69 65 4 0.02 2.22 1950 89 85 4 0.02 1.94 1930 73 65 8 0.042 4.16 1950 72 69 3 0.01 1.39 1946 96 89 7 0.04 3.87 1926 81 73 8 0.042 4.16 1942 90 72 18 0.09 9.29 1943 97 96 1 0.00 0.40 1919 83 81 2 0.008 0.83 1934 97 90 7 0.04 3.74 1925 98 97 2 0.01 0.83 1918 97 83 14 0.075 7.48 1932 98 97 2 0.01 0.83 1918 111 98 12 0.07 6.57 1917 98 97 2 0.008 0.83 1918 114 98 16 0.08 8.32 1917 114 111 3 0.02 1.74 1904 114 98 16 0.083 8.32 1917 165 114 51 0.27 27.23 1901 176 114 62 0.33 32.70 1901 130 114 16 0.083 8.32 1901 189 165 24 0.13 12.68 1850 189 176 14 0.07 7.22 1839 189 130 60 0.316 31.60 1885 189 0 0.000 0.00 1826 189 0 0.00 0.00 1826 189 0 0.000 0.00 1826 Max. 51 0.272 27.2 Max. 62 0.327 32.7 Max. 60 0.316 31.6 Total 1.00 100 Total 1.00 100 Total 1.00 100

* More examples for recruitment of Moringa peregrina are available online: http://dx.doi.org/10.17632/z23sx6vjc3.138TU U38T

95

3.4 Relictic populations of the endangered tree Moringa peregrina (Moringaceae) in Sinai, Egypt are still genetically diverse Mohamed Dadamouny*, Anja Klahr, Franziska Eichhorn, and Martin Schnittler Ernst-Moritz-Arndt University of Greifswald, Institute of Botany and Landscape Ecology, Soldmannstr. 15, D-17487 Greifswald, Fax +49 03834 420 4114 *Email: [email protected]

Short title: Genetic diversity of Moringa peregrina revealed with microsatellite markers

Abstract: We studied genetic diversity of relictic populations of the endangered tree Moringa peregrina at Sinai Peninsula, Egypt. Microsatellite markers from M. oleifera were used to analyse 259 samples (63%) of 407 mapped trees in four wadis, in addition to 148 (57%) repeats. Twelve of 13 markers and 88.5±2.2% of all loci were polymorphic. For each second tree, the analysis was repeated, resulting in 2% error per multilocus genotype. Six clones, each comprising two trees, were found, with its members always occurring within one wadi and the older tree located upstream. This points to vegetative dispersal via branches uprooted after flash floods. AMOVA showed 83% of the total genetic variance to occur within populations. Observed and expected heterozygosities for the total population were 0.423 and 0.469, respectively; eight loci deviated significantly from HWE. The FST of 0.133 overall loci indicated little differentiation between the four populations. STRUCTURE, using figures for K between 1 and 7, revealed the most pronounced genetic structure for K=3. Two neighboured wadis (W. Agala and W. Feiran) were weakly differentiated; the most remote wadi (Zaghra) had as well the most deviating genotypes. Some outliers indicated occasional gene flow, most likely due to the use of seeds by Bedouins. Our results fit into a pattern of relictic populations isolated not very long ago. The preserved diversity on one hand, but the rapid decline of all populations, on the other hand, indicates a good chance for success of a conservation programme but underlines as well its urgency. Keywords: Diversity, heterozygosity, genetic structure, Multiplex PCR, Moringa, Sinai, SSR 1. Introduction The angiosperm family Moringaceae with its 13 species is restricted to the dry tropics (Quattrochi 2000; Dadamouny et al. 2016). One species, M. oleifera, is widely cultivated, and microsatellite markers (Vieira et al. 2016) have been developed (Wu et al. 2010; Selvakumari and Ponnuswami, 2015). In this study, we report the cross-amplification and application of these markers to estimate the genetic diversity and the degree of isolation for relictic populations of a second species, M. peregrina, at the Sinai Peninsula, Egypt. M. peregrina (Forssk.) Fiori is a medicinal tree growing on the Arabian Peninsula and the horn of Africa. The remaining populations of M. peregrina in Sinai are increasingly threatened by climate change (Dadamouny and Schnittler 2016) and over-collection for medicinal uses, leading to poor recruitment (Zaghloul et al. 2012b, Dadamouny et al. 2016). In a rapidly changing environment due to natural reasons (increasing aridity after the mid-Holocene climate optimum in Saharan Africa 9000–6000 years ago, Clausen et al. 2003; Kröpelin et al. 2008) that are now acting synergistic with current, most likely anthropogenic, changes in climate (Dadamouny and Schnittler 2016), genetic diversity of the few remaining populations might be of fundamental importance to ensure survival of the tree. Establishing conservation programs could help in mitigating the adverse impacts of these threats, but require data on genetic diversity and degree of isolation to set the right focus. Therefore, we aimed to study genetic diversity, heterozygosity and gene flow between four populations of M. peregrina, to detect signs of possible genetic erosion due to inbreeding. Most of the few genetic studies in Moringaceae (see Yadav and Srivastava 2014 for review) were directed towards M. oleifera as the most common and economically important species (e.g., Muluvi et al. 1999; Wu et al. 2010; Shahzad et al. 96 Genetic diversity of Moringa peregrina revealed with microsatellite markers Publications

2013; Ganesan et al. 2014; Natarajan and Joshi 2015; Rajalakshmi et al. 2017). Only three studies targeted M. peregrina, using allozymes (Zaghloul et al. 2012a), Inter-Simple Sequence Repeats (ISSRs, Al Khateeb et al. 2013), and Internal Transcribed Spacer (ITS, Alaklabi (2015). However, in a preliminary study, we did not found any polymorphism for ITS for 28 samples of M. peregrina from Sinai (GenBank numbers MH153752–MH153779). The fourth study including both species (M. peregrina and M. oleifera) used RAPD and ISSR Markers (Robiansyah et al. 2015). According to the phylogenetic study of Olson (2002a) for Moringaceae, M. oleifera and M. peregrina are closely related to each other, and some authors lump the two taxa in spite of distinct geographical and ecological differences (Olson 2002a, b). Since RAPD markers, which were used in Moringa (Ojuederie et al. 2013; Saini et al. 2013; Smit et al. 2013; Popoola 2014) suffer from low reproducibility, we tested SSR markers from M. oleifera for its applicability in M. peregrina.

2. Material and Methods 2.1 Study area and sampling The mostly mountainous Sinai Peninsula supports a rather diverse vegetation and the first traces of human inhabitants’ data back about 7000 years (Danin 1983; Zohary 1973, 1983, 1999; Zahran and Willis 1992, 2009). During regular surveys at the Sinai Peninsula from 2007 to 2016, we mapped the remaining four populations of M. peregrina at four wadis (Fig. 1), with different histories of disturbance and fragmentation (see Dadamouny 2009; Dadamouny et al. 2012, 2016). These wadis are 1. Agala (WA, with a total of 35 trees) which is a short and narrow tributary of a greater wadi (W. Feiran). It is about 4 km long and 50–100 m wide. 2. Wadi Feiran (WF, with a total of 45 trees); it is one of the longest wadis in Southern Sinai (about 30 km long and 400–500 m wide). 3. Wadi Meir (WM, with a total of 242 trees) lies north-east of El-Tor. It is about 18 km long; the flat wadi bed is up to 200–300 m wide and dissected by channels and deltas due to the water erosion in rainy years, as it is subjected to many floods, recently in 2010 (Dadamouny and Schnittler 2016). 4. Wadi Zaghra (WZ, with a total of 71 trees) is situated northeast of Saint Catherine near Dahab city (≈21 km), and ca. 40 km long and about 300–500 m wide. We collected data on tree diameter, tree vitality, and tree age based on the correlation of tree diameter to a number of annual rings (see Dadamouny et al. 2016). In summer 2014, we sampled the silica gel-dried leaves of 407 trees (including 12 new saplings). From these 407 M. peregrina trees, 259 samples were selected for this study by a grid-based approach which has been shown to reflect a maximum of genetic diversity with a minimum of sampling (Suzuki et al. 2005) for DNA extraction and genotyping. 2.2 DNA Extraction, measuring, and dilution Three methods for DNA extraction were tried to deal with the high content of phenolics in M. peregrina leaves. First, the CTAB extraction protocol of Doyle and Doyle (1987) and (1990) was adapted, including more steps of washing with ammonium acetate (3–5 times) in order to obtain a clear supernatant that yielded a good amount of DNA in 1 µl (Fig. S1). A second extraction protocol included an extraction Robot (KingFisher Flex) using Omega M1130 extraction Kits. The third extraction protocol employed Invisorb spin columns and 96-plates using Stratec kits (Stratec molecular, Germany) following the manufacturer’s protocol. Moringa leaves yielded a high amount of DNA for all three tested extraction methods: CTAB (1000–3000 ng), Stratec kit (300–1000 ng), and Robot (20–200 ng), although many samples failed for the latter. Due to its good time/cost /quality ratio, the Stratec kit was routinely used. We measured each sample by Nanodrop (Thermo-Fisher Scientific, USA) at least three times to calculate the average of DNA concentration, and then we diluted it to 100 ng, 50 ng and 25 ng. 2.3 Optimizing the PCR protocol From 20 genomic microsatellite primers developed for M. oleifera (Wu et al. 2010), we tested 16 SSR primers, of these, 13 worked well with the DNA from M. peregrina (Table S1). Protocols for PCR reactions were modified slightly from Ganesan et al. (2014) and Shahzad et al. (2013). The first test with non-fluorescent primers was carried out in a volume of 25 µl. We used the following master mix: 15.44 µl ddH2O, 2.5 µl 10x Axon-Buffer (Axon scientific, Malaysia),

3.4 µl MgCl2 (Axon, 25 mM), 0.5 µl dNTP (10 mM), 0.16 µl AxonTaq (5 U/µl), 1 µl of each primer F and R (5 or 10

97

Publications Genetic diversity of Moringa peregrina revealed with microsatellite markers pmol/µl) combined with 1 µl of 25– 0 ng DNA per sample. Different annealing temperatures for each primer were tested in a gradient PCR and the following conditions were found best: initial denaturation at 95 oC for 5 min, 94 oC for 1 min, 35 cycles of 1 min for primer annealing temperatures of 55–60 oC, 1 minute at 72 oC; and 8 minutes at 72 oC for the final extension, holding at 4oC. To check the obtained SSR motifs, for each primer PCR products were purified and sequenced. First, we did the DNA purification using 1.1 vol of PEG, then 5 ng of purified DNA used for cycle sequencing PCR which was done using the forward primer only, 0.25 µl Big Dye and 0.25 µl Half Big Dye. At that time we purified the PCR products of the cycle sequencing then sequenced by capillary ABI sequencer (Applied Biosystems). The obtained SSR motifs were compared with those of M. oleifera (Wu et al. 2010) (examples are shown in Fig. S2 and Table S2). 2.4 Establishing a Multiplex PCR assay for M. peregrina With 13 primers (tested with Auto-dimer software, Vallone and Butler, 2004) combined with three multiplex PCRs, the genetic diversity of 259 samples of M. peregrina was investigated. A uniform annealing temperature of 60 oC was found to achieve the best amplification (Table S1). For routine use, multiplex PCRs were scaled down to 6 µl, including 3 µl master mix (see below), 1 µl from primer combination for each group, and 1-2 µl of diluted DNA (25–50 ng), and 1 µl

H2O. Two multiplex master mix kits were tested (KAPA2G Fast Multiplex; Kapa Biosystems; Multiplex PCR kits, Quiagen) with different concentrations of DNA (100 ng, 50 ng, 25 ng, 10 ng per 1 µl) to optimise primer concentrations for the multiplex reactions (Table S1), with the KAPA2G kit performing best. Products were analyzed in a plate of 96- wells using an ABI Prism 3100 Genetic Sequencer (Applied Biosystems) at the Institute of Zoology, University of Greifswald, Germany. 2.5 Fragment analysis and marker evaluation We scored the alleles with GeneMapper® software version 5. And optimizing the binning of raw data with Tandem software (Matschiner and Salzburger 2009, see Fig. S3). A total of 148 repeats for 259 plants (57%) are used to calculate the error rate. These samples were repeatedly extracted and genotyped under the same conditions, as recommended by Hoffman and Amos (2005). Different error rates were calculated (Table S3): per-allele by dividing the number of differently typed alleles by the total number of re-typed alleles; per locus by dividing the number of re-typed loci with at least one mismatching allele by the total number of re-typed loci; per multilocus genotype (MLG) by dividing the number of repeated genotypes including one mismatch by the total number of repeated genotypes (Pompanon et al. 2005). All electropherograms were checked at least three times in GeneMapper to minimize the risk of false scoring. Furthermore, the dataset was checked for errors and null alleles with the micro-checker software (Van Oosterhout et al. 2004) and null allele frequency was also assessed using Genepop v 4.7 (Rousset, 2008). For check more the marker resolution, a genotyping accumulation curve for the 259 genotyped samples were drawn with R package Poppr (version 2.2.0, Kamvar, Brooks and Grünwald 2015). The genotype curve was built by randomly sampling x loci and counting the number of observed MLGs through the “gac” and “poppr” function (Nei 1978; Pielou 1975; Grünwald et al. 2003; Kamvar and Grünwald, 2014). To assess the resolution of the assay, this procedure was repeated 1000 times for 1 locus up to n-1 loci, creating n-1 distributions of observed MLGs. 2.6 Statistical analyses To evaluate the genetic diversity between the four populations, we calculated several population genetic parameters (Fixation index, test for Hardy-Weinberg equilibrium, alleles frequencies, AMOVA, etc.) using the programs GenALEx 6.5 (Peakall and Smouse 2012), Genepop (Raymond and Rousset 1995; Rousset 2008), and the R packages PopGenReport and Poppr. According to Wright (1978), the ranges of FST indicate the genetic differentiation to be 0.0 – 0.05: little

(genotypes are closed to each other and high gene flow), 0.05 – 0.15: moderate, 0.15 – 0.25: great, FST above 0.25: very great genetic differentiation. As diversity measures incorporate both genotypic richness and abundance, to evaluate the genetic diversity between the populations, we calculated Heterozygosity, evenness (as a measure of the distribution of genotype abundances). Further, we employed STRUCTURE (ver. 23.4, July 2012) to analyze the population structure, in order to infer the historical lineages that show clusters of similar genotypes and to infer gene flow using multilocus

98

Genetic diversity of Moringa peregrina revealed with microsatellite markers Publications genotype data (Pritchard et al. 2000; Falush, Stephens and Pritchard 2003). The implemented model assigns each individual to one assumed population (K values from 1 to 7 were tested), determining the assumed number of populations according to on maximum likelihood (LnPD) or delta K (Δk) values (Evanno et al. 2005). We also used the program STRUCTURE HARVESTER on the web (Earl 2012) for 168 runs (Burn-in periods of 50000 and 250000 iterations, and the sampling phase of 500000 repeats) to carry out downstream processing of STRUCTURE results to calculate Evanno’s Δk value. For an analysis of the similarity and dissimilarity of each population, data are evaluated with MicrosatAllele, and PopGenReport, then we did PCA using the R routines NumXl. The populations were clustered also based on dissimilarity analysis using DARwin software (version 6, Perrier and Jacquemoud-Collet 2006). Clones were identified by direct pairwise comparison of all MLGs; using thresholds for errors of 0, 1, and 2 deviating loci per MLGs (Arnaud-Hanod et al. 2007). This resulted in 6, 9, and 22 clones, but for thresholds >0 we would have obtained 1 and 2 clones, respectively, with trees in different wadis (other than WA and WV, see Fig. 3). 3. Results 3.1 Microsatellite assay Figure 1 shows the distribution of these 259 trees with known genotype in relation to the number of non-analyzed (148) and missed (not found in 2014 compared with 2006; 31) trees. Fig. S1 shows the obtained band under different annealing temperatures. SSR motifs in the studied populations of M. peregrina were comparable with those of M. oleifera but varied in length (Fig. S2, Table S2). We processed 407 samples (including also 148 repeats to determine the error rate, see Table 1) representing 259 out of 407 mapped trees of M. peregrina from four wadis (Fig. 1). All 13 markers except MO41 are informative. Table S3 shows details on the error rate (0.16% per locus, 2% per MLG). The range of allele sizes, the transformed values by the software Tandem (10 scorings for locus MO6 and 6 scorings for locus MO10 corrected), and the frequency of the alleles for the 13 investigated loci are shown in Fig. S3 and Table S1. Null alleles were recorded for 13 samples (3.19 %), all belonging to locus MO10. Checking peak positions of all analyzed fragments, an average peak position error of 0.12 bp by Tandem was found. The best performing locus was MO46 (0.06 bp), the worst MO13 (0.16 bp).

Fig. 1. a–b Location of the study area and map of Sinai Peninsula with the populations of M. peregrina (A) at four wadis (black dots), e–f Maps showing locations investigated trees for Wadi Feiran (b), W. Zaghra (c), W. Meir (d), and W. Agala (e); coordinates are given in the UTM system (grid 36R). Green circles represent trees analyzed in this study (N), orange circles living trees not sampled (n), red circles indicate dead or uprooted trees (m). Ten individuals germinating after 2006 were analyzed (blue dots, symbol Ne). One outlier tree for Wadi Zaghra is not shown on the map (coordinates 633129 E and 3170736 N).

99

Publications Genetic diversity of Moringa peregrina revealed with microsatellite markers

Table 1. Observed (HO) and expected (HE) heterozygosities and allelic richness (figures below the abbreviations for the wadis) of the populations (133 samples from W. Meir WM, 29 from W. Agala WA, 39 from W. Feiran WF, and 58 from W. Zaghra WZ). Shown are as well the total number of alleles recorded for each locus (#All.), intra-individuals (Qi) and inter-individuals intra-population (Qii) values and F-Statistics over all populations for each locus. Locus HO HE WM WA WF WZ Locus #All. Qi Qii FIS FIT FST MO1 0.125 0.126 1.34 1.00 1.00 2.00 MO1 2 0.112 0.119 0.006 0.348 0.345 MO6 0.402 0.456 3.37 3.61 2.99 3.39 MO6 7 0.432 0.476 0.119 0.314 0.221 MO10 0.069 0.195 3.48 2.75 3.58 4.27 MO1 10 0.077 0.168 0.648 0.654 0.017 MO55 0.506 0.515 5.87 4.00 5.08 5.64 MO50 10 0.474 0.487 0.016 0.070 0.054 MO61 0.332 0.333 2.19 2.70 2.00 2.40 MO65 5 0.397 0.394 0.004 0.113 0.109 MO41 0.000 0.000 1.00 1.00 1.00 1.00 MO41 1 0.000 0.000 Monomorphic MO45 0.621 0.610 5.34 5.58 3.92 5.33 MO41 10 0.617 0.632 -0.018 0.121 0.137 MO46 0.665 0.621 5.48 5.88 6.97 5.53 MO45 8 0.672 0.642 -0.071 0.024 0.088 MO56 0.581 0.652 5.26 3.98 6.33 5.80 MO56 13 0.579 0.634 0.109 0.210 0.113 MO8 0.446 0.458 3.41 4.68 3.56 4.27 MO06 6 0.378 0.392 0.026 0.223 0.202 MO12 0.300 0.705 7.14 5.31 6.39 8.66 MO18 12 0.317 0.727 0.574 0.596 0.050 MO13 0.713 0.697 4.52 5.61 6.54 5.36 MO12 9 0.694 0.688 -0.023 0.125 0.145 MO68 0.742 0.730 6.37 4.98 6.17 5.45 MO63 12 0.745 0.725 -0.017 0.100 0.115 Mean 0.423 0.469 4.21 3.93 4.27 4.54 Mean8 8.1 0.423 0.468 0.115 0.241 0.133 Mean allelic richness per tree 5.6 1.1 1.7 2.6 SE 0.0 0.068 0.067 0.066 0.056 0.025

When the loci were sampled 1,000 times without replacement, a genotype accumulation curve demonstrated that 6–7 markers were necessary to discriminate between 90% of the multilocus genotypes and at least 7–8 markers were necessary to discriminate between 100% of the multilocus genotypes (Fig. 2). One locus (MO41) was monomorphic without exception, a second (MO1) was monomorphic for the two most closely related wadis (WA and WF). The highest percentage of polymorphic loci (92.3 %) was recorded for WM and WZ. At the same time, the two closest wadis (WF and WA) exhibited the same percentage of polymorphic loci (84.6 %).

Fig. 2. Genotype accumulation curve of 259 Moringa peregrina for 13 loci. Box plots show the number of multilocus genotypes for 407 repeats with loci randomly sampled without replacement (253 multilocus genotypes and 6 clones). The red dashed line represents 100% of the total observed multilocus genotypes. The blue line has been drawn using the R functions ggplot2 and method 'gam'.

3.2 Allelic richness and F-Statistics of population All but one marker (MO41) were polymorphic and revealed altogether 105 alleles (without repeated samples). Table 1 shows the allelic richness per locus, ranging from 1 (MO41) to 8.66 (MO12). From the four populations, the highest allelic richness was found for samples of W. Zaghra (9) at MO12. However, average allelic richness per tree for a wadi was highest in W. Meir WM (5.6), followed by the trees of W. Zaghra WZ (2.6), Feiran WF (1.7) and the trees at W. Agala WA (1.1). Observed (HO) and expected heterozygosity (HE) ranged from 0 to 0.86 and from 0 to 0.81 for each wadi (Table S4). The loci MO45, MO46, MO13, and MO68 exhibited the highest heterozygosity (observed, expected and unbiased) and the highest differentiation intra-individuals and inter-individuals intra-population (Table 1). For the total population, HO and HE ranged from 0.069 to 0.742 and from 0.126 to 0.730 with averages of 0.423 and 0.469, respectively (Table 1). 100

Genetic diversity of Moringa peregrina revealed with microsatellite markers Publications

Table 1 shows as well FIT, which can be partitioned into FST (mean 0.133) due to the Wahlund effect and FIS (mean 0.115) due to inbreeding (individuals in a population are more related than to expect by chance). Percentages of polymorphic loci were 92.3%, 84.6%, 84.6%, and 92.3% for the wadis WM, WA, WF, and WZ, respectively (average 88.5±2.2%). Table S4 shows for each wadi, population size, number of alleles (N), number of effective alleles (Na), information index (I), observed heterozygosity (HO), expected (HE) and unbiased expected heterozygosity (uHe) per locus are shown. The probability of identity (PI) for all 13 loci was 2.58 x 10-9 (Fig. S4). Micro-checker revealed that except for MO10 there was no evidence for scoring errors due to stuttering or for large allele dropout, and no evidence for null alleles. GenAlEx outputs showed that MO13 and MO68 are exhibiting the highest differentiation between individuals. All 13 loci deviated from HWE, and this was significant (p<0.001) for 8 loci, probably due to a considerable degree of inbreeding (Table S5), although the degree of polymorphy of the markers varied considerably. For all genetic diversity indices (e.g. heterozygosity index, Stoddart and Taylor’s Index, expected heterozygosity), the trees of the most populous W. Meir (133 of 242 trees analyzed) performed best, followed by the trees of W. Zaghra (58 of 71 trees analyzed), which includes the oldest trees at the Sinai (Fig. S5). However, genotypic evenness (Evenness E.5) as the ratio of the number of abundant genotypes to the number of rarer genotypes was slightly better for the W, Agala with the lowest number of trees population followed by the W. Feiran, Zaghra, and Meir. Another pairwise population matrix of Nei Genetic Distance and identity and totals for allelic patterns by populations are shown in Table S7. An analysis of molecular variance (AMOVA) revealed 16% genetic variation to occur between populations, 9% among the individuals and 75% within individuals (83% within the population if within individuals’ analysis was suppressed). The mean values for allelic patterns and its standard error by population are shown in Table S8. 3.3 Gene flow between populations Table S6 shows Nei genetic distance, Nei genetic identity, and the total values for Allelic Patterns by populations. Figure S6 shows the mean Allelic Patterns across the populations. Based on the G-statistics, figures for HE and HO range from 0.126 to 0.730 (average of 0.469) and from 0.125 to 0.742 (average of 0.423), respectively. However, the expected heterozygosity (HE) and observed heterozygosity (HO) ranged from 0.126 to 0.730 (average of 0.469) and from 0.125 to 0.742 (average of 0.423). The population assignment output by GenAlEx to 'Self' or 'Other' Population (with leaving out one option) revealed that 91% of the populations were assigned to “self” (likely to be sired from trees within the population) and 9% to “others”. The numbers of self/others were 124/9 in W. Meir (WM), 24/5 in W. Agala (WA), 35/4 in W. Feiran (WF), and 53/5 in W. Zaghra. The assignment indicated as well that WA and WF are genetically close to each other, followed by WA and WM. FST and GST values are shown in Table 2. All populations are genetically differentiated as the

FST value is higher than 0.05, however, the second and the third population (WA and WF) are close to each other (FST = 0.0542). Trees of WZ are most deviating from the trees of other wadis. Although private allelic richness was low, private alleles were recorded for 27 individuals of 133 (20 %) in WA and 22 of 58 (38 %) in WZ.

Table 2. FST (lower left) and GST values (upper right) based on the variance of allele frequencies, with the number of private alleles (#PA), of individuals possessing private alleles (#IndPA), and the probability of identity (PI) for the four wadis (left).

Pairwise Population FST/GST Values #PA #IndPA PI WM WA WF WZ WM 0.000 0.048 0.043 0.135 WM 18 27 9.5×10-08 WA 0.095 0.000 0.028 0.115 WA 4 4 6.2×10-08 WF 0.087 0.054 0.000 0.112 WF 2 1 1.1×10-08 WZ 0.237 0.210 0.203 0.000 WZ 11 22 5.0×10-07 WM WA WF WZ PI for all 2.6×10-09

3.4 Genetic Structure of the M. peregrina populations We used the model implemented in STRUCTURE to determine the genetic relationship among genotypes of M. peregrina trees. It was assumed that the number of populations was K and the loci were independent and at Hardy– 101

Publications Genetic diversity of Moringa peregrina revealed with microsatellite markers

Weinberg equilibrium. A range from K = 1 to K = 7 was tested, Ln(PD) kept on increasing with increasing population number but there was no clear population structure, therefore Ln(PD) derived k was plotted against the K to determine the number of populations using the software “Structure harvester” available online. At K = 3 a maximum figure for K was found (Fig. 3) and this was considered as the true number of M. peregrina populations for 259 genotypes. Figure S8 shows the values of Evanno’s Δk und different Ks. In addition, we obtained the same trend in a principal component analysis (Fig. 4). The trees are grouped according to alleles (fragment sizes) into four groups. The largest represents the trees of W. Meir, the second and the third group (W. Agala and W. Feiran) are close together; genetically most isolated appears the fourth group (W. Zaghra). A few outliers may be explained by translocation of seeds between wadis by Bedouins and, maybe, former research activities. Factorial analysis based on dissimilarity in alleles’ sizes (Fig. S6A) and the dendrogram of the Unweighted Pair Group Method with Arithmetic Mean (Fig. S6B) confirms as well this pattern. Most similar to each other are again trees numbered 1 to 133 from W. Meir (green) and trees 134 to 162 from W. Agala (red). In accordance, a neighbour-joining tree based on dissimilarity analysis for M. peregrina (Fig. S7) assigns nearly all trees from W. Zaghra to a separate clade and unites the three other populations to a second clade, although the internal structure of this clade is as well influenced by population differences. With a few exceptions, trees from one wadi cluster together, illustrating the low gene flow between populations. Again, trees of the second and the third wadi (WA and WF) are genetically closest to each other.

Fig. 3. Genetic Structure of the M. peregrina populations assuming different K values (a, b: K =2, c, d: K= 3, and e, f: K = 4). Colors denote the assignment of samples to populations, shown in a simplified diagram featuring the geographic distances (lines) and extent of sampling (size of circles) for the wadis. g F-Statistics visualized for the four populations: line thickness scaled according to F-value (thus a thinner line indicates higher gene flow). h Plot of Ln versus K obtained with Structure harvester. 102

Genetic diversity of Moringa peregrina revealed with microsatellite markers Publications

Fig. 4. Principal Component Analysis (PCA) for markers combination based on SSR data for 259 M. peregrina samples collected from the four wadis; each dispersion ellipse marker a shape is a cluster for the trees of one wadi; Wadi Meir WM (green), Wadi Agala WA (red), Wadi Feiran WF (blue), and W. Zaghra WZ (yellow).

3.5 Clonal growth in M. peregrina trees Six pairs of trees were identical for all 13 loci and represent putative clones (Table 3), with trees of a clone always occurring in the same wadi, and the larger tree upstream. Figure S9 shows the combinations of samples and the number of matching loci for clonal analysis in M. peregrina trees with all analyzed trees (259) and genotypes (253).

Table 3. List of six clones recorded in three wadis; showing their location, distances between trees (m), the tree diameter at breast height (dbh), and estimated age. The larger of the two trees for a clone is listed first.

UTM Tree Code Wadi Easting Northing Distance [m] dbh [cm] Age [yrs] 57---- 314--- S8T3 WM 3727 8987 8 34.2 67 S8T5_N WM 3719 8985 1.3 3 S14T3 WM 4451 9626 2 30.6 81 S14T2 WM 4449 9625 30.4 80 S18T2 WM 5053 9560 6 141.1 360 S18T3 WM 5058 9563 41.2 109 S21T12 WM 5582 9381 92 23.2 61 S21T2 WM 5504 9429 13.7 40 S26T12 WA 836 4098 284 20.7 59 S26T6 WA 818 4381 15.9 46 S33T12 WZ 6547 780 174 32.6 163 S34T4 WZ 6694 874 60.2 153

4. Discussion

SSR assay Moringa is the single genus of a family Moringaceae with thirteen species (Olson and Carlquist 2001). Therefore, cross-amplification of the primers designed by Wu et al. (2010) for M. oleifera to the most closely related species M. peregrina (Olson 2002b, El-Awady et al. 2015) seemed prospective, since motifs and flanking regions are often conserved across related taxa (Rossetto 2001) and associated with nonrepetitive DNA (Morgante et al. 2002). As not uncommon for desert plants, the extraction of highly purified DNA for M. peregrina was difficult due to high contents of sugars and phenolic compounds. ITS sequences for 28 M. peregrina trees from all four wadis in Sinai were identical, and the same was found by Alaklabi (2015) for 14 samples from Saudi-Arabia. However, the latter study showed differences between ITS genotypes for trees from different locations in Saudi Arabia, which is in contrast to results from Al-Sobeaia, Alamer, and Amer (2015) who found that M. peregrina and M. oleifera cannot be distinguished by ITS sequences - which lends evidence that the two species are closely related. This was shown as well for M. oleifera in India (GenBank numbers KT737744– KT737801) by Boopathi and Kumar (2015). Therefore it is no surprise that 13 of the 16 tested SSR primers 103

Publications Genetic diversity of Moringa peregrina revealed with microsatellite markers from oleifera worked for peregrina. The 13 loci used for this study reliably amplified, but only 11 loci were sufficiently polymorphic. The low error rate indicates that SSR is suitable for genotyping M. peregrina. We found recorded only 13 samples null alleles (3.19%, all in locus MO10), which are commonly caused by sequence variations at the primer target site that will prevent primer annealing and will result in no amplification (Dakin and Avise 2004; Chapuis and Estoup 2007). The presence of null alleles in SSR assays may cause an underestimation of heterozygosity (Porth and El-Kassaby 2014). A heterozygous individual with one null allele is usually scored as homozygous at that specific locus (Dąbrowski et al. 2014), which has obvious implications for population and parentage studies. In addition, SSR markers are known to show a high mutation rate (Mason, 2015). Spatial genetic structure fits best into a pattern of relictic populations The multilocus genotypes revealed a clear differentiation between populations, which correlates well with geographic distance between them. The remaining populations seem still to preserve a considerable degree of genetic diversity. We could not find data for FST values for other trees in the area. Wu et al. (2010) stated that the expected (HE) and observed (HO) heterozygosities for the proposed twenty SSR primers in M. oleifera (material from Bangalore in India and Pakokku in Myanmar) ranged from 0.36 to 0.76 (average of 0.55) and from 0.00 to 0.87 (average of 0.46), respectively. Our study, covering a much smaller region, yielded lower figures for M. peregrina, with HE ranging from 0.13 to 0.73 (average of 0.47) and HO from 0.12 to 0.74 (average of 0.42), respectively. In addition, deviations from HWE were detected in all 13 loci and were significant for 10 loci (p<0.001) and for 8 loci in Chi-Square Test (p<0.001). Given the rather high seed weight (Abd El-Wahab 1995, Dadamouny 2009) and the lack of special dispersal mechanisms, the four remaining populations appear to be now reproductively isolated. This may explain the deviations from HWE (found in all loci, significant for eight of them, P < 0.001), which are most likely is caused by a considerable level of inbreeding (FIS is ranging from -0.07 to 0.65). However, the effective number of alleles Ae and allelic richness Ar (see Duwe et al. 2017), was highest for the geographically most isolated (but largest) population from W. Zaghra (WZ), followed by Wadi Meir (WM) and W. Feiran (WF). The population with the lowest number of trees showed as well the lowest allelic richness (W.

Agala, WA). FST values exceeding 0.05 (average 0.133) indicate a moderate genetic differentiation for the four populations. The populations of the second and the third wadi (WA and WF) are geographically close (distance ca. 2 km) and exhibit as well the lowest FST (0.054). GST values (Table 2) were comparable, although for highly polymorphic loci GST by itself provides virtually no information about the actual degree of differentiation of subpopulations and is not monotonic with respect to increasing differentiation (Ryman and Leimar 2009). The spatial genetic structure is underpinned by figures for trees exhibiting private alleles, recorded for 27 of 133 individuals of WA (20 %) but 22 of 58 in WZ (38 %).

In accordance with these figures, the analysis with STRUCTURE revealed that K=3 to fit best the existing genetic structure, since one wadi (W. Agala) is a tributary of a second wadi (W. Feiran), and thus likely to be accessible via gene flow, maybe with seed dispersal by flash floods (Dadamouny and Schnittler 2016). This pattern is explained best by the assumption that the remaining four populations of M. peregrina at the Sinai are the leftovers of a formerly contiguous population, and this appears likely if looking at the climatic history of the region. Saharan Africa experienced a humid phase between 9000–6000 BC (Claussen et al. 2003) and progressive drying from this time onwards (Kröpelin et al. 2008). This seems to be similar at the Arab peninsula, where remainders of former lakes reveal cyclic changes with an earlier humid phase (35–20 000 BC, Preusser 2009). As such, we can assume that the tree is somewhat adapted to climate changes.

However, the last few decades brought a downward trend in annual rainfall in all of eastern Africa, which is well known for the Sahel zone (lasting since ca. 1950, Nicholson 1989), for Saudi-Arabia (Almazroui et al. 2012) and was detected by us as well for the Sinai (Dadamouny and Schnittler 2016). Together with human overuse of the medicinal plant, this is most likely the main reason for the rapid decline (Dadamouny et al. 2016) of all populations. This, most likely anthropogenic, acceleration of climate change seems now to pose a real threat for the tree, as it is indicated by signs of genetic erosion for the smallest population (W. Agala).

104

Genetic diversity of Moringa peregrina revealed with microsatellite markers Publications

Clonal propagation may hardly facilitate survival in M. peregrina As a novelty, a very few investigated trees proved to be clonal (identical in all 13 loci). Trees of a clone occurred always in the same wadi, and the estimated age of the tree located upstream was higher than that of the downstream tree (Table 3). This supports the assumption that flash floods, which often destroy trees (Zaghloul et al. 2012b; Dadamouny and Schnittler 2016), may carry branches downstream which can root again. Flash floods are not uncommon in the region (five to seven drought years, then a heavy rainy year). However, this affects only trees at the wadi beds, which show an overall better performance than those at the slopes of the wadis, but is much more endangered by flash-floods (Dadamouny et al. 2016). This trade-off between growth and disturbance lets clonal propagation appear more an exit strategy than a regular mode of reproduction, although it might prove to be very valuable for conservation. Propagation by cuttings worked well in experiments carried out for M. peregrina (from W. Feiran: Abd El-Wahab 1995; W. Feiran and W. Zaghra: Dadamouny 2009; Dadamouny et al. 2012). Moringa shares cloning induced by unfavourable conditions with the genus Populus (Dunlap and Stettler 1996; Bradshaw et al. 2000), which occurs like M. peregrina in river valleys (Rood and Mahoney, 1990). In difference to Moringa, root suckering is a regular phenomenon in poplars (Bradshaw et al. 2000), whereas trees of M. peregrina tend to become multi-stemmed but do not develop horizontal roots, a precondition for successful propagation by root suckering (Wiehle et al. 2009). Our study revealed the lion’s share of trees to originate by sexual propagation, having clearly distinct multilocus genotypes. Pollinators of M. peregrina trees are most likely bees like Anthophora pauperata (Gilbert et al. 1996), Xylocopa sp., Coelioxys sp., flies like Eristalinus taeniops, Lucilia sericata, Eupeodes corollae, butterflies, and other wadi insects (although wind cannot be ruled out). Dadamouny (2009) and Dadamouny et al. (2016) studied flowering phenology, but from the Sinai Peninsula, no data on pollination are available. Our results suggest that a substantial gene flow occurred only between the trees of the large W. Feiran and W. Agala, which is a short and narrow tributary of the former. A pronounced spatial genetic structure is not uncommon for rare species with a few populations only; very similar results were found for the rare species Erythrina sandwicensis, which is endemic to the Hawaiian Islands (Young et al. 1996; Münster and Wieczorek 2007). What can be done to ensure the survival of M. peregrina at the Sinai? Although the four remaining populations of M. peregrina from the Sinai appeared still surprisingly diverse, SSR studies for the related M. oleifera from Pakistan and nine other countries of the world (Shahzad et al. 2013; Ganesan et al. 2014) revealed 6-13 alleles per locus (in contrast to 2-13 for M. peregrina) and a higher level of observed heterozygosity in the accessions collected from Pakistan as compared to the accessions from the nine different countries (maintained at ECHO, USA). As one of the few tree species and virtually the only tree that can be directly used by humans, M. peregrina is a kind of keystone species for the vegetation of the Sinai. We thus agree with Shahzad (2013) that the conservation of M. peregrina is not only important to maintain biodiversity, but is desirable as well from an ethnobotanical, dietary and pharmaceutical perspective. As it appears now, the genetic consequences as a result of the habitat fragmentation (see Young et al. 1996) seem to be a less severe than the missing recruitment and population decline. The remaining M. peregrina populations in Sinai may face a possible genetic erosion more from low recruitment (Dadamouny et al. 2016) and human activities of cutting and seeds collection, i.e. loss of individuals than from inbreeding. According to Hollingsworth et al. (2005), circa situ conservation and sustainable use strategies (Simons et al. 1994; Weber et al. 2001) should be considered.

We recommend the following actions to ensure the survival of M. peregrina at the Sinai: (1) Conserving the genetic resources for the remaining populations, giving priority to the most diverse population of W. Meir, plus the most deviating in W. Zaghra. Trees of these two wadis have as well the highest annual growth (Dadamouny et al. 2016). This may include in-situ and ex-situ conservation programs, like propagation of trees and planting of young individuals at the margins of wadi beds, where survival changes are highest. (2) Collecting seeds from the trees and applying paternity analysis (which can be achieved with the marker set presented in this study) together with pollination experiments. This will help to understand the mating system of M. peregrina trees, 105

Publications Genetic diversity of Moringa peregrina revealed with microsatellite markers especially if they are obligatory or facultative allogamous. In turn, it answers the question if assisted pollination helps to preserve genetic diversity. (3) In parallel, selection and subsequent breeding of trees to optimize traits like pod weight and/or oil content in seeds (see Souza et al. 2013) may facilitate acceptance for such a program by local inhabitants. In the current situation, the species requires assisted propagation to persist, best by artificial cross-breeding of trees from different wadis to have genetically diverse offspring and a subsequent reforestation program. Acknowledgements The first author appreciates the support by the German Academic Exchange Service (DAAD) and Ministry of Higher Education, Egypt in the frame of the German-Egyptian Long-term Scholarship program (GERLS). In addition, we would like to thank the team of St. Catherine Protectorate of for their efforts in protecting the endangered species of M. peregrina in the area and facilitating the fieldwork. David Würth helped with the evaluation of microsatellite data using GeneMapper; Manuela Bog provided useful comments on the manuscript. Supplementary data Supplementary figures, tables and datasets are available online: http://dx.doi.org/10.17632/ytphxbtzhk.1 References Abd El-Wahab RH (1995) Reproduction ecology of wild trees and shrubs in Southern Sinai, Egypt. M.Sc. Thesis, Bot. Department, Faculty of Sci, Suez Canal University, Ismailia, Egypt. p.110 Al-Khateeb W, Bahar E, Lahham J, Schroeder D, Hussein E (2013) Regeneration and assessment of genetic fidelity of the endangered tree Moringa peregrina (Forsk.) Fiori using Inter-Simple Sequence Repeat (ISSR). Phys Mol Biol Plants 19(1):157–164. https://doi.org/10.1007/s12298-012-0149-z Alaklabi, A (2015) Genetic diversity of Moringa peregrina species in Saudi Arabia studied with ITS sequences. Saudi J Biol Sci 22(2):186–190. https://doi.org/10.1016/j.sjbs.2014.09.015 Al-Sobeai SM, Alamer KH, Amer S (2015) Molecular identification of the cultivated Moringa oleifera in Saudi Arabia. Int J Plant Animal Env Sci 5(3): 28–32 Almazroui M, Nazrul M, Athar H, Jones PD, Ashfaqur Rahman M (2012) Recent climate change in the Arabian Peninsula: annual rainfall and temperature analysis of Saudi Arabia for 1978–2009. Int J Climatology 32(6): 953–966. Arnaud-Hanod S, Duarte CM, Alberto F, Serrão EA (2007) Standardizing methods to address clonality in population studies. Mol Ecol 16:5115–5139. https://doi.10.1111/j.1365-294X.2007.03535.x Bradshaw HD, Ceulemans R, Davis J, Stettler R (2000) Emerging model systems in plant biology: poplar (Populus) as a model forest tree. J Plant Grow Regul 19(3):306–313. https://doi.org/10.1007/s003440000030. Boopathi MN, Kumar S (2015) Utility of ITS region for DNA barcoding and phylogenetic analysis of Moringa (Moringa oleifera Lam.) germplasm. https://www.ncbi.nlm.nih.gov/nuccore/KT737744.1 Accessed December 2017. Chapuis MP, Estoup A (2007) Microsatellite null alleles and estimation of population differentiation. Mol Biol Evol 24(3):621– 631. Claussen M, Brovkin V, Ganopolski A, Kubatzki C, Petoukhov V (2003) Climate change in northern Africa: The past is not the future. Climatic Change 57(1-2):99–118 Dąbrowski MJ, Pilot M, Kruczyk M, Żmihorski M, Umer HM, and Gliwicz J (2014) Reliability assessment of null allele detection: inconsistencies between and within different methods. Mol Ecol Res 14(2):361–373. https://doi.org/10.1111/1755-0998.12177 Dadamouny MA (2009) Population ecology of Moringa peregrina growing in Southern Sinai, Egypt. M.Sc. thesis, Suez Canal University, Faculty of Science, Ismailia. 205 pp. Dadamouny MA, Zaghloul MS, Moustafa, AA (2012). Impact of improved soil properties on establishment of Moringa peregrina seedlings and a trial to decrease its mortality rate. Egy J. Boty 52(1):83–98. Dadamouny MA, Schnittler M (2016) Trends of climate with rapid change in Sinai, Egypt. J Water Climate Change 7(2):393– 414. https://doi.org/10.2166/wcc.2015.215 Dadamouny MA, Unterseher M, König P, Schnittler M (2016) Population performance of Moringa peregrina (Forssk.) Fiori (Moringaceae) at the Sinai Peninsula, Egypt in the last decades: consequences for its conservation. J Nat Conserv 34, 65– 74 doi. 10.1016/j.jnc.2016.08 Dakin EE, Avise JC (2004) Microsatellite null alleles in parentage analysis. Heredity 93(5):504–509. https://doi.org/10.1038/sj.hdy.6800545 Danin A (1983) Desert Vegetation of Israel and Sinai. Jerusalem, Cana Publishing House. 148 pp. Doyle JJ, Doyle JL (1987) A rapid DNA isolation procedure from small quantities of fresh leaf tissues. Phytochem Bull 19:11– 15 Doyle JJ, Doyle JL (1990) Isolation of plant DNA from fresh tissue. Focus 12:13–15. Dunlap JM, Stettler RF (1996) Genetic variation and productivity of Populus trichocarpa and its hybrids. IX. Phenology and Melampsora rust incidence of native black cottonwood clones from four river valleys in Washington. Forest Ecol Manag 87(1-3):233–256. https://doi.org/10.1016/S0378-1127(96)03774-7

106

Genetic diversity of Moringa peregrina revealed with microsatellite markers Publications

Duwe VK, Muller LAH, Reichel K, Borsch T, Ismail SA (2017) Genetic structure and genetic diversity of the endangered grassland plant Crepis mollis (Jacq.) Asch. as a basis for conservation management in Germany. Conserv Genet https://doi.org/10.1007/s10592-017-1025-8 Earl, DA (2012) STRUCTURE HARVESTER: a website and program for visualizing STRUCTURE output and implementing the Evanno method. Conserv Genet Res 4(2):359–361. https://doi.org/10.1007/s12686-011-9548-7 El-Awady MA, Hassan MM, Abdel-Hameed ESS, Gaber A (2015) Comparison of the antimicrobial activities of the leaves-crude extracts of Moringa peregrina and Moringa oleifera in Saudi Arabia. Int J Curr Microbiol App Sci 4(12):1–9. Evanno G, Regnaut S, Goudet J (2005) Detecting the number of clusters of individuals using the software STRUCTURE: a simulation study. Mol Ecol 14(8):(2611–2620. https://doi.org/10.1111/j.1365-294X.2005.02553.x Falush D, Stephens M, Pritchard JK (2003) Inference of population structure using multilocus genotype data: linked loci and correlated allele frequencies. Genetics 164(4):1567–1587. Ganesan KS, Singh R, Choudhury DR, Bharadwaj J, Gupta V, Singode A (2014) Genetic diversity and population structure study of drumstick (Moringa oleifera Lam.) using morphological and SSR markers. Indust Crop Prod 60:316–325. https://doi.org/10.1016/j.indcrop.2014.06.033 Gilbert F, Willmer P, Semida F, Ghazoul J, Zalat S (1996) Spatial variation in selection in a plant-pollinator system in the wadis of Sinai, Egypt. Oecologia 108(3):479–487 Grünwald NJ, Goodwin SB, Milgroom MG, Fry WE (2003) Analysis of genotypic diversity data for populations of microorganisms. Phytopathol 93:738–746. https://doi.org/10.1094/PHYTO.2003.93.6.738 Hoffman JI, Amos W (2005) Microsatellite genotyping errors: detection approaches, common sources and consequences for paternal exclusion. Mol Ecol (2):599–612 Hollingsworth PM, Dawson IK, Goodall-Copestake WP, Richardson JE, Weber JC, Sotelo Motes C, Pennington RT (2005) Do farmers reduce genetic diversity when they domesticate tropical trees? A case study from Amazonia. Mol Ecol 14(2):497– 501. https://doi.org/10.1111/j.1365-294X.2005.02431.x Kamvar ZN, Grunwald NJ (2014) Poppr 1.0.3: an R package for genetic analysis of populations with mixed (clonal/sexual) reproduction. https://doi.org/10.7717/peerj.281 Kamvar ZN, Brooks JC, Grünwald, NJ (2015) Novel R tools for analysis of genome-wide population genetic data with emphasis on clonality. Front Genet 6:208. https://doi.org/10.3389/fgene.2015.00208 Kröpelin S, Verschuren D, Lézine A-M, Eggermont H, Cocquyt C, Francus P, Cazet J-P, Fagot M, Rumes B, Russell JM, Darius F, Conley DJ, Schuster M, von Suchodoletz H, Engstrom DR (2008) Climate-driven ecosystem succession in the Sahara: The past 6000 years. Science 320:765–768 Mason AS (2015) SSR Genotyping. In: Batley J (Ed) Plant Genotyping. Springer, New York, NY, pp 77– 89. https://doi.org/10.1007/978-1-4939-1966-6_6 Matschiner M, Salzburger W (2009) TANDEM: Integrating automated allele binning into genetics and genomics workflows. Bioinformatics 25(15):1982–1983. ISSN 13674803. https://doi.org/10.1093/bioinformatics/btp303. Morgante M, Hanafey M, Powell W (2002) Microsatellites are preferentially associated with nonrepetitive DNA in plant genomes. Nat Genet 30:194–200. https://doi.org/10.1038/ng822 Muluvi GM, Sprent JI, Soranzo N, Provan J, Odee D, Folkard G, McNicol JW, Powell W (1999) Amplified fragment length polymorphism (AFLP) analysis of genetic variation in Moringa oleifera Lam. Mol Ecol 8(3):463–470. https://doi.org/10.1046/j.1365-294X.1999.00589.x Münster P, Wieczorek AM (2007) Potential gene flow from agricultural crops to native plant relatives in the Hawaiian Islands. Agriculture, Ecosyst and Env 119(1–2):1–10. https://doi.org/10.1016/j.agee.2006.06.014 Natarajan S, Joshi JA (2015) Characterisation of Moringa (Moringa oleifera Lam.) Genotypes for growth, pod and seed characters and seed oil using morphological and molecular markers. Int J Plant Res 28(2):64–71 Nei M (1978) Estimation of average heterozygosity and genetic distance from a small number of individuals. Genetics 89(3):583– 590 Nicholson SE (1989) Long-term changes in African rainfall. RMetS Weather 44(2):46–56 Ojuederie OB, Igwe DO, Okuofu SI, Faloye B (2013) Assessment of genetic diversity in some Moringa oleifera Lam. landraces from western Nigeria using RAPD markers. Afr J Plant Sci Biotechnol 7(1):15–20 Olson ME (2002a) Intergeneric relationships within the Caricaceae-Moringaceae clade (Brassicales) and potential morphological synapomorphies of the clade and its families. Int J Plant Sci 163:51–65. https://doi.org/10.1086/324046. Olson ME (2002b) Combining data from DNA sequences and morphology for a phylogeny of Moringaceae (Brassicales). System Bot 27:55–73 Olson ME, Carlquist S (2001) Stem and root anatomical correlations with life form diversity, ecology, and Systematics in Moringa (Moringaceae). Bot J Linn Soc 135:315–348. https://doi.org/10.1111/j.1095-8339.2001.tb00786.x. Peakall R, Smouse PE (2012) GenAlEx 6.5: genetic analysis in Excel. Population genetic software for teaching and research – an update. Bioinformatics 28, 2537–2539. https://doi.org/10.1093/bioinformatics/bts460. Perrier X, Jacquemoud-Collet JP (2006) DARwin software http://darwin.cirad.fr/ Accessed December 2017 Pielou EC (1975) Ecological diversity. John Wiley & Sons, New York. Pompanon F, Bonin A, Bellemain E, Taberlet P (2005) Genotyping errors: causes consequences, and solutions. Nat Rev Genet 6:847–859. https://doi.org/10.1038/nrg1707 Popoola JO (2014) Genetic diversity in Moringa oleifera from Nigeria using fruit morphometric characters and random amplified polymorphic DNA (RAPD) markers. Covenant J Phys Life Sci 1(2):43–50. Porth I, El-Kassaby, YA (2014) Assessment of the genetic diversity in forest tree populations using molecular markers. Diversity 6(2):283–295. https://doi.org/10.3390/d6020283 Preusser F (2009) Chronology of the impact of Quaternary climate change on continental environments in the Arabian Peninsula. Comptes Rendus Geoscience 341(8-9):621–632 107

Publications Genetic diversity of Moringa peregrina revealed with microsatellite markers

Pritchard, JK, Stephens M, Donnelly P (2000) Inference of population structure using multilocus genotype data. Genetics 155:945–959 Quattrochi U (2000) CRC World dictionary of plant names: Common names, scientific names, eponyms, synonyms, and etymology. Vol 3, CRC Press, Washington. 887 pp Rajalakshmi R, Rajalakshmi S, Parida A (2017). Evaluation of the genetic diversity and population structure in drumstick (Moringa oleifera L.) using SSR markers. Cur Sci 112(6):1250. https://doi.org/10.18520/cs/v112/i06/1250-1256. Raymond M, Rousset F (1995) An exact test for population differentiation. Evol 49(6):1280–1283. https://doi.org/10.1111/j.1558-5646.1995.tb04456.x Robiansyah I, Ramadan A, Al-kordy MA, Ghushash AS, Hajar AS (2015) Characterization of a new biotype Moringa of Saudi Arabia using RAPD and ISSR markers. Bulletin Kebun Raya 18(2):99–110. Rood SB, Mahoney JM (1990) Collapse of riparian poplar forests downstream from dams in western prairies: probable causes and prospects for mitigation. Env Manag 14(4):451–464. https://doi.org/10.1007/BF02394134 Rossetto M (2001) Sourcing of SSR markers from related plant species. In: Henry R (ed) Plant genotyping – the DNA fingerprinting of plants. CAB Int, Wallingford UK, pp 211–224. https://doi.org/10.1079/9780851995151.0211 Rousset F (2008) Genepop: a complete reimplementation of the Genepop software for 193 Windows and Linux. Mol. Ecol. Res. 8, 103–106. doi. 10.1111/j.1471-1948286.2007.01931.x Ryman N, Leimar O (2009) GST is still a useful measure of genetic differentiation–a comment on Jost's D. Mol Ecol 18:2084– 2087. https://doi.org/10.1111/j.1365-294X.2009.04187.x. Saini RK, Saad KR, Ravishankar GA, Giridhar P, and Shetty NP (2013) Genetic diversity of commercially grown Moringa oleifera Lam. cultivars from India by RAPD, ISSR and cytochrome P450-based markers. Plant Syst Evol 299(7):1205–1213. https://doi.org/10.1007/s00606-013-0789-7 Selvakumari P, Ponnuswami V (2015) Genetic diversity of Moringa (Moringa oleifera Lam.) using SSR markers. Int. J Trop Agriculture 33(2):943–946. Shahzad U (2013) Micro cloning and morphogenetic analysis of Moringa species. Dissertation, University of Agriculture, Faisalabad. Shahzad U, Khan MA, Jaskani MJ, Khan IA, Korban SS (2013). Genetic diversity and population structure of Moringa oleifera. Cons Genet 14(6):1161–1172. https://doi.org/10.1007/s10592-013-0503-x Simons AJ, MacQueen DJ, Stewart JL (1994) Strategic concepts in the domestication of non-industrial trees. In: Leakey RRB, Newton AC (eds) Tropical Trees: The Potential for Domestication and the Rebuilding of Forest Resources HMSO, London, UK, pp 91–102 Smit R, Du Toit ES, and Vorster, BJ (2013) RAPD and SSR genetic diversity analysis of Moringa oleifera. S Afr J Bot (86):182. https://doi.org/10.1016/j.sajb.2013.02.162 Souza LM, Gazaffi R, Mantello CC, Silva CC, Garcia D, Le Guen V, Cardoso SEA, Garcia AAF, and Souza AP (2013) QTL Mapping of growth-related traits in a full-sib family of rubber tree (Hevea brasiliensis) evaluated in a sub-tropical climate. PLoS One 8(4): e61238. https://doi.org/10.1371/journal.pone.0061238 Suzuki J, Herben, T, and Maki, M (2005) An under-appreciated difficulty: sampling of plant populations for analysis using molecular markers. Evol Ecol 18:625–646. https://doi.org/10.1007/s10682-004-5147-3 Vallone PM and Butler, JM (2004) AutoDimer: a screening tool for primer-dimer and hairpin structures. Biotechniques 37(2): 226–231. https://doi.org/10.2144/04372ST03 Van Oosterhout C, Hutchinson WF, Wills DP, Shipley P (2004) MICRO‐CHECKER: software for identifying and correcting genotyping errors in microsatellite data. Mol Ecol Res 4(3):535–538. https://doi.org/10.1111/j.1471-8286.2004.00684.x Vieira MLC, Santini, L, Diniz, AL, Munhoz, CDF (2016) Microsatellite markers: what they mean and why they are so useful. Genet Mol Biol 39(3):312–328 Weber JC, Sotelo Montes C, Vidaurre H, Dawson IK, Simons AJ (2001) Participatory domestication of agroforestry trees: an example from the Peruvian Amazon. Development in Practice 11: 425–433. https://doi.org/10.1080/09614520120066710 Wiehle M, Eusemann, P, Thevs N, Schnittler M (2009) Root suckering patterns in Populus euphratica (Euphrates poplar, Salicaceae). Trees 23(5):991–1001. https://doi.org/10.1007/s00468-009-0341-0 Wright S (1978) Evolution and the Genetics of Populations. Vol 4. Variability Within and Among Natural Populations. Univ. of Chicago Press, Chicago. Wu JC, Yang, J, Gu, ZJ, and Zhang, YP (2010. Isolation and characterization of twenty polymorphic microsatellite loci for Moringa oleifera (Moringaceae). Hort Sci 45(4):690–692 Yadav S, Srivastava J (2014) Genetic diversity analysis of Moringa oleifera by using different molekular markers: A review. Int Journal Recent Scientific Research 5(12):2277–2282 Young A, Boyle T, Brown T 1996) The population genetic consequences of habitat fragmentation for plants. Trends Ecol Evol 11(10): 413–418. https://doi.org/10.1016/0169-5347(96)10045-8 Zaghloul MS, Hamrick JL, Moustafa AA (2012a) Conservation genetics of Sinai’s remnant populations of Moringa peregrina, an economically valuable medicinal plant. Conserv Genet 13(1):9–19. https://doi.org/10.1007/s10592-011-0260-7 Zaghloul MS, Moustafa AA, Dadamouny MA (2012b) Age structure and population dynamics of Moringa peregrina, an economically valuable medicinal plant. Egypt J Bot 52(2):499–519 Zohary M (1973) Geobotanical foundations of the Middle East. Geobotanica Selecta, volume 3. Gustav Fischer Verlag, 739 pp. Zohary M (1983) Man and vegetation in the Middle East. Geobotany. In: Holzner W, Werger MJA, Ikusima I (Eds) Man's impact on vegetation. Dr W. Junk BV Publishers, The Hague, pp 287–295 Zohary D (1999) Monophyletic vs. polyphyletic origin of the crops on which agriculture was founded in the Near East. Genetic Resources and Crop Evolution, 46(2):133–142 Zahran MA, Willis AJ (1992) The vegetation of Egypt. London, Chapman and Hall, 424 pp Zahran MA, Willis AJ (2009) The vegetation of Egypt. 2nd Ed, London, Chapman and Hall, 437 pp 108

Appendix 07

Appendix VII

Relictic populations of the endangered tree Moringa peregrina (Moringaceae) in Sinai, Egypt are still genetically diverse

Fig. S1. Gel profile for the DNA Extraction using CTAB buffer (a), SSR bands run in 3 % agarose gel for 1 hr at 100 V; 10 of 11 SSR primers produced clear band (b) and the reaction was optimized under different annealing temperature (c and d).

Fig. S2. Examples of sequences of purified PCR products reveal the SSR motifs; homozygous CT 18x (A), and heterozygous GA 11x (B) for M. peregrina trees at Sinai, Egypt. 109 Appendix 07

Full range

A

Frequency

B

C

D

E

F

Allele size in pb (observed / transformed)

Fig. S3. Tandem plot for the binning of the 13 loci reveals the range of Fig. S4. Probability of Identity (PI) for each Locus for each population allele sizes (black) and the transformed values (grey) and the frequency of (A-D), for pooled population (E), and for increasing combinations of the alleles. Low frequencies in the red circle seem to be outliers. the 13 Loci (F).

110

Appendix 07

Fig. S5. Summary of genetic diversity showing H: heterozygosity Index, G: Stoddart and Taylor’s Index, Hexp: Expected Heterozygosity (Nei, 1978), and E.5 Evenness; which is the ratio of the number of abundant genotypes to the number of rarer genotypes calculated by R

(library: "poppr") for the four wadis (Meir WM, Feiran WF, Zaghra WZ, and the total). The coloured dots refers to the each observed

Fig. S6. (A) factorial analysis based on dissimilarity index (simple matching) calculated by DARwin Software for 259 M. peregrina samples collected from the four wadis; Wadi Meir WM (green), Wadi Agala WA (red), Wadi Feiran WF (blue), and W. Zaghra WZ (yellow) represent the genotypes from different populations in four clusters. (B) UPGMA dendrogram generated from Nei's genetic distance on the four populations.

Fig. S7. Unweighted Neighbour-Joining Tree based on dissimilarity analysis calculated by DARwin Software for the studied M. peregrina populations at the four wadis; W. Meir (green), W. Agala (red), W. Feiran (blue), and W. Zaghra (yellow) . 111 Appendix 07

Fig. S8. Delta K for M. peregrina populations under different K calculated by Structure Harvester shows a sharp decrease in delta-K values from K2 to K3, followed by a plateau at K4.

Fig. S9. Pairwise combinations of multilocus genotypes the analzed M. peregrina trees; (A) without repeats, where the number of genotypes was 253 and (B) with repeats, where the number of genotypes was 258. Bars in the histogram show the number of combinations identical in 1, 2,…,13 loci.

112

Genetic diversity of Moringa peregrina revealed with microsatellite markers

Relictic populations of the endangered tree Moringa peregrina (Moringaceae) in Sinai, Egypt suffer are still genetically diverse

Table S1. List of selected primers; sequence, motifs, allele size and the tested annealing temperature for each primer in a singleplex PCR and finally 60 oC that we recommended as an annealing temperature for the three groups of the multiplex. The length of amplicon and intensity of the band. The optimum concentration and colour for the multiplex PCR, and binning by software (average of rounding error and number of modified samples out of 407 sample). Tandem binning Singleplex PCR Multiplex PCR (407 samples) Prim GeneBank Pair Locus Primer sequence Repeat motif Ann. Avg. # accession # Allele size Final Tested length of Band Final conc. No. mean Temp Color G Rounding modified (bp) Ann. Temp. amplicon (mM) (oC) Intens error indiv. F: TTGTCTGCCTCCTTTTGTCA GQ853887 ++ 3 MO1 AACTGTCACCCTCCTATCCA (TC)17 150–158 155 58 58.1 200 Y 0.010 red 1 0.13 0

GQ853888 F: GCATAGCCACCTTTACTCCT ++++ 2 MO6 R: TAGTGGGTCCAAGACAAAGC (AG)7-(AG)6 497–520 510 61 60.3 400 Y 0.014 blue 1 0.13 10 F: CTTTACACCTCAGTATCCCT ++++ 6 MO10 GQ853890 R: GTTCGGCTTATGTTCTCGTT (CT)20-(CT)5 294–306 300 58 58.1 300 Y 0.016 blue 1 0.13 6 F: ATTACAGAACGATGAAACCA GQ853900 ++ 16 MO55 R: CTCTTTCCCTCCATTCAACC (AG)8AA(AG)14 110–114 115 56 55.7 120 Y 0.070 yellow 1 0.11 0 F: TGTGGGTCCTGCCTTTTCTC GQ853903 ++++ 8 MO61 R: CTTCTGTCTTTCTTCCTGCT (TC)11 296–300 300 60 60.3 340 Y 0.014 green 1 0.11 0 F: TGGGATTAGGGCATTAGAAA +++ 7 MO41 GQ853895 R: TAGTGGGTCCAAGACAAAGC (GA)10 147–153 150 55 55.7 150 Y 0.011 yellow 2 0.10 0 F: CCTTTGAAGTTGAAAATCTC +++ 1 MO45 GQ853897 R: TTCTAGGGTAGTTGAATCCA (TC)5TT(TC)10 200–208 205 58 58.1 220 Y 0.031 red 2 0.10 0 F: ACCAAGGGTTTCAACTGCTG ++++ 14 MO46 GQ853898 R: CATTTTGCGACGGTCTCACG (AG)5-(GA)6 420–440 430 61 60.3 350 Y 0.020 green 2 0.06 0 F: TCAATACGCCAAGTAAGCAA ++ 5 MO56 GQ853901 R: AAGCACTTCACGCATAAAAC (AG)13 346–350 350 60 60.3 320 Y 0.044 blue 2 0.11 0 F: GTAGATGGTGCAGCTACTCA ++ 11 MO8 GQ853889 R: TGGGGTTCTTGTTCTTTATT (CT)13 150–156 150 58 58.1 190 Y 0.011 green 3 0.11 0 GQ853891 F: ACCGAAGATGATAAGGTGGG ++++ 4 MO12 R: CAAAAGGAAGAACGCAAGAG (CT)11 260–270 265 59 58.9 290 Y 0.014 red 3 0.12 0 F: TTTCGGGTTTTCTTTCACGG ++++ 9 MO13 GQ853892 R: AGCTCACTTTCCATCTCCAT (CT)15 331–341 335 58 58.1 330 Y 0.010 blue 3 0.16 0 GQ853906 F: TGCTTCGCTTCCTCTATTCT +++++ 13 MO68 R: ACCACAGGCTTGCTTCAGTA (GA)12 240–250 245 56 55.7 370 Y 0.050 yellow 3 0.13 0

Table S2 Variations in motif length between Moringa peregrina obtained from the tree leaflets grown in Sinai, Egypt and Moringa oleifera obtained from germinated seeds collected in Bangalore (India) and Pakokku (Myanmar).

SSR Motifs M. peregrina M. oleifera* SSR Motifs M. peregrina M. oleifera* AG 5x 13x TC 11x 11x, 17x CT 5x, 7x, 18x 11,13,15x GA 5x, 11x 10,12,13x *Moringa oleifera data are from Wu et al. 2010. HortSci. 45(4), 690–692.

113 Appendix 08 114 Appendix 08 Table. S3. The error rate for the obtained Table S4 Sample size, number of alleles (N), number of effective alleles (Na), information index (I), observed heterozygosity (Ho), expected (He) and multilocus genotypes. unbiased expected Heterozygosity (uHe), and fixation index (F) for each population for 259 M. peregrina samples collected from the four wadis; Wadi Meir WM, Wadi Agala WA, Wadi Feiran WF, and W. Zaghra WZ.

Genotyping Errors # % Pop MO1 MO6 MO10 MO55 MO61 MO41 MO45 MO46 MO56 MO08 MO12 MO13 MO68

WM N 133 133 133 133 133 133 133 133 133 133 133 133 133 No. of markers 13 Na 2 5 7 10 3 1 7 7 10 5 11 6 10 No. of alleles/marker 2 Ne 1.015 1.999 1.140 1.854 2.010 1.000 3.622 3.632 2.419 1.290 3.776 3.365 3.477 I 0.044 0.900 0.342 1.013 0.714 0.000 1.451 1.455 1.167 0.489 1.604 1.334 1.493 No. of repeats 148 Ho 0.000 0.481 0.083 0.451 0.526 0.000 0.707 0.737 0.564 0.211 0.278 0.722 0.759 He 0.015 0.500 0.123 0.461 0.503 0.000 0.724 0.725 0.587 0.225 0.735 0.703 0.712 No. of singles 259 uHe 0.015 0.502 0.123 0.462 0.504 0.000 0.727 0.727 0.589 0.226 0.738 0.705 0.715 F 1.000 0.037 0.327 0.020 -0.047 #N/A 0.024 -0.017 0.039 0.065 0.622 -0.027 -0.066 Total no. of sample runs 407 WA N 29 29 26 29 29 29 29 29 29 29 29 29 29 Na 1 4 3 4 3 1 6 6 4 5 6 6 5 No. of allelic mismatches: 4 Ne 1.000 2.064 1.215 2.588 1.550 1.000 3.398 2.484 2.700 1.687 3.132 4.313 4.014 I 0.000 0.858 0.365 1.117 0.593 0.000 1.386 1.266 1.139 0.860 1.302 1.539 1.472 No. of replicated alleles: 3848 Ho 0.000 0.414 0.038 0.621 0.345 0.000 0.862 0.690 0.517 0.379 0.207 0.828 0.724 He 0.000 0.515 0.177 0.614 0.355 0.000 0.706 0.598 0.630 0.407 0.681 0.768 0.751 uHe 0.000 0.525 0.180 0.624 0.361 0.000 0.718 0.608 0.641 0.414 0.693 0.782 0.764 Mean error rate per allele: no of allelic mismatches/no of 0.001 0.10 F #N/A 0.197 0.782 -0.012 0.028 #N/A -0.222 -0.154 0.178 0.069 0.696 -0.077 0.036 replicated alleles WF N 39 39 35 39 39 39 39 39 39 39 39 39 39 Na 1 3 4 6 2 1 4 7 7 4 8 7 7 No. of SLG with at least 1 3 Ne 1.000 1.486 1.429 2.315 1.195 1.000 3.107 4.791 3.915 2.245 3.172 5.290 4.467 mismatch: I 0.000 0.612 0.610 1.129 0.302 0.000 1.199 1.740 1.563 0.961 1.385 1.753 1.596 No. of replicated SLG: 1924 Ho 0.000 0.282 0.057 0.538 0.179 0.000 0.641 0.821 0.641 0.538 0.128 0.821 0.795 He 0.000 0.327 0.300 0.568 0.163 0.000 0.678 0.791 0.745 0.555 0.685 0.811 0.776 Mean error rate per locus: no. of SLG with 1 uHe 0.000 0.331 0.304 0.575 0.166 0.000 0.687 0.802 0.754 0.562 0.694 0.822 0.786 0.002 0.16 mismatch/no. of replicated F #N/A 0.138 0.810 0.052 -0.099 #N/A 0.055 -0.037 0.139 0.029 0.813 -0.012 -0.024 SLG WZ N 58 58 52 58 58 58 58 58 58 58 58 58 58 No. of MLG with at least 1 Na 2 4 6 7 3 1 8 7 8 5 10 7 6 3 mismatch: Ne 1.953 1.936 1.221 1.712 1.450 1.000 1.501 1.592 2.841 2.816 3.550 2.023 3.128 I 0.681 0.849 0.461 0.913 0.521 0.000 0.780 0.849 1.276 1.173 1.668 1.038 1.323 No. of replicated MLG: 148 Ho 0.500 0.431 0.096 0.414 0.276 0.000 0.276 0.414 0.603 0.655 0.586 0.483 0.690 He 0.488 0.484 0.181 0.416 0.310 0.000 0.334 0.372 0.648 0.645 0.718 0.506 0.680 uHe 0.492 0.488 0.183 0.419 0.313 0.000 0.337 0.375 0.654 0.651 0.725 0.510 0.686 Mean error rate per MLG 0.020 2.03 F -0.025 0.109 0.468 0.005 0.112 #N/A 0.173 -0.112 0.069 -0.016 0.184 0.045 -0.014

Genetic diversity of Moringa peregrina revealed with microsatellite markers Genetic diversity of Moringa peregrina revealed with microsatellite markers

Table S5. Summary of Chi-Square Tests for Hardy-Weinberg Equilibrium; ns=not significant, * P<0.05, ** P<0.01, *** P<0.001 Pooled population WM WA WF WZ Locus DF ChiSq Prob Signif DF ChiSq Prob Signif DF ChiSq Prob Signif DF ChiSq Prob Signif DF ChiSq Prob Signif MO1 1 35.318 0.000 *** 1 133.000 0.000 *** Monomorphic Monomorphic 1 0.035 0.851 ns MO6 21 43.601 0.003 ** 10 5.783 0.833 ns 6 4.812 0.568 ns 3 5.522 0.137 ns 6 11.481 0.075 ns MO10 45 956.698 0.000 *** 21 266.260 0.000 *** 3 26.012 0.000 *** 6 59.721 0.000 *** 15 109.872 0.000 *** MO55 45 44.301 0.501 ns 45 9.632 1.000 ns 6 7.707 0.260 ns 15 14.276 0.505 ns 21 21.456 0.431 ns MO61 10 7.419 0.685 ns 3 1.422 0.700 ns 3 3.899 0.273 ns 1 0.379 0.538 ns 3 1.126 0.771 ns MO41 Monomorphic Monomorphic Monomorphic Monomorphic Monomorphic MO45 45 48.991 0.316 ns 21 11.574 0.950 ns 15 51.304 0.000 *** 6 7.518 0.276 ns 28 70.324 0.000 *** MO46 28 28.913 0.417 ns 21 16.656 0.732 ns 15 11.571 0.711 ns 21 20.024 0.520 ns 21 41.522 0.005 ** MO56 78 255.816 0.000 *** 45 79.669 0.001 ** 6 5.779 0.448 ns 21 100.888 0.000 *** 28 98.327 0.000 *** MO08 15 76.427 0.000 *** 10 13.863 0.179 ns 10 9.605 0.476 ns 6 4.423 0.620 ns 10 41.227 0.000 *** MO12 66 590.394 0.000 *** 55 289.875 0.000 *** 15 64.332 0.000 *** 28 128.598 0.000 *** 45 195.587 0.000 *** MO13 36 178.276 0.000 *** 15 19.248 0.203 ns 15 11.677 0.703 ns 21 28.459 0.128 ns 21 54.742 0.000 *** MO68 66 134.382 0.000 *** 45 68.239 0.014 * 10 16.611 0.083 ns 21 23.032 0.342 ns 15 58.473 0.000 ***

Table S6. Global Hardy-Weinberg tests [Score (U) test] and Hardy-Weinberg test (FIS estimates) for 259 M. peregrina samples collected from the four wadis; Wadi Meir WM, Wadi Agala WA, Wadi Feiran WF, and W. Zaghra WZ using 13 SSR markers. Global Hardy-Weinberg tests [Score (U) test] MO1 MO6 MO10 MO55

P-val S.E. switches P-val S.E. W&C R&H Swit P-val S.E. W&C R&H Swit. P-val S.E. W&C R&H Swit P-val S.E. W&C R&H Swit (ave.) WM 0.0000 0.0000 10417.1 WM 0.0037 0.0003 1.0000 1.0076 358 0.2450 0.0244 0.0410 0.0184 11890 0.0000 0.0000 0.3299 0.3310 1011 0.2237 0.0197 0.0242 0.0151 3056 WA 0.0000 0.0000 16759.9 WA - 0.1457 0.0114 0.2140 0.0809 12464 0.0007 0.0004 0.7899 0.5191 4900 0.2031 0.0079 0.0059 0.0873 29844 WF 0.0000 0.0000 18816.0 WF - 0.0690 0.0021 0.1504 0.2287 28418 0.0000 0.0000 0.8145 0.6312 7886 0.6124 0.0181 0.0650 -0.019 8686 WZ 0.0000 0.0000 14858.6 WZ 0.6477 0.0040 -0.016 0.0161 78735 0.0275 0.0000 0.0000 0.2287 28418 0.0000 0.0000 0.4758 0.3378 1273 0.3199 0.0227 0.0137 0.0064 4501 Tot. 0.0000 0.0000 15100.9 Global Hardy-Weinberg MO61 MO45 MO46 MO56 Loci tests [Score (U) test] MO1 0.0025 0.0003 39605.0 P-val S.E. W&C R&H Steps P-val S.E. W&C R&H Swit P-val S.E. W&C R&H Steps P-val S.E. W&C R&H Swit MO6 0.0125 0.0026 17487.8 WM 0.6901 0.0220 -0.043 -0.022 29708 0.3067 0.0306 0.0275 0.0138 17323 0.5192 0.0323 -0.013 -0.008 19340 0.0169 0.0062 0.0426 0.0765 3281 MO10 0.0000 0.0000 3663.3 WA 0.4142 0.0079 0.0460 0.0009 15564 0.4788 0.0115 -0.205 0.0009 12753 0.9636 0.0040 -0.137 -0.081 12364 0.1258 0.0046 0.1954 0.1074 30321 MO55 0. 2374 0.0268 11639.0 WF 1.0000 0.0000 -0.085 -0.086 25263 0.4356 0.0077 0.0677 0.0073 36144 0.3977 0.0154 -0.024 0.0144 23611 0.0106 0.0026 0.1518 0.1728 14461 MO61 0.5174 0.0179 21671.5 WZ 0.2506 0.0077 0.1201 0.0670 15331 0.0108 0.0037 0.1817 0.0834 1832 0.9687 0.0117 0.1037 0.0337 3871 0.0028 0.0019 0.0775 0.1396 5216 MO41 - - - MO8 MO12 MO13 MO68 MO45 0.0569 0.0073 16977.3 MO46 0.7829 0.0231 14715.0 P-val S.E. W&C R&H Steps P-val S.E. W&C R&H Swit P-val S.E. W&C R&H Steps P-val S.E. W&C R&H Swit MO56 0.0000 0.0000 13310.5 WM 0.2303 0.0146 0.0683 0.0203 5972 0.0000 0.0000 0.6239 0.2462 6711 0.8041 0.0223 0.0232 0.0255 20072 0.9482 0.0180 0.0623 0.0333 6820 MO08 0.1139 0.0076 13303.8 WA 0.1527 0.0092 0.0861 0.0873 7612 0.0000 0.0000 0.7050 0.5100 9881 0.7640 0.0120 0.0599 0.0478 17768 0.0518 0.0043 0.0531 0.1495 31912 MO12 0.0000 0.0000 7384.25 WF 0.5909 0.0173 0.0420 0.0162 20604 0.0000 0.0000 0.8171 0.5155 5197 0.2019 0.0108 0.0012 0.0494 21861 0.4626 0.0227 0.0112 0.0066 13879 MO13 0.5610 0.0154 16244.8 WZ 0.2225 0.0108 0.0072 0.0340 19787 0.0000 0.0000 0.1923 0.3958 7477 0.2792 0.0219 0.0539 0.0187 5766 0.0024 0.0008 0.0051 0.2322 16921 MO68 0.0844 0.0132 17461.3

115 Appendix 08 116 Appendix 08 Table S7. Pairwise population matrix of Nei Genetic Distance (left), Nei genetic identity (middle), and totals for allelic patterns by populations. Na = No. of different alleles, Na (Freq >= 5%) = No. of different alleles with a frequency more than or equal 5%, Ne = No. of effective alleles, No. LComm Alleles (<=25%) = No. of locally common alleles (Freq. >= 25% and 50%) found in 25% or fewer populations (which was 0) and found in 50% or fewer populations. SSR data for 259 M. peregrina samples collected from the four wadis; Wadi Meir WM, Wadi Agala WA, Wadi Feiran WF, and W. Zaghra WZ.

WM WA WF WZ WM WA WF WZ Population WM WA WF WZ 0.000 WM 1.000 Na 84 54 6 74 0.107 0.000 WA 0.898 1.000 Na Freq. >= 5% 38 41 45 33 0.096 0.074 0.000 WF 0.908 0.929 1.000 Ne 18 4 2 11 0.310 0.272 0.267 0.000 WZ 0.734 0.762 0.765 1.000 No. LComm Alleles (<=25%) 0 0 0 0 WM WA WF WZ WM WA WF WZ No. LComm Alleles (<=50%) 16 17 17 16

Table S8. Mean values and standard error for allelic patterns by populations for 259 M. peregrina samples collected from the four wadis; Wadi Meir WM, Wadi Agala WA, Wadi Feiran WF, and W. Zaghra WZ, where Na = No. of Different Alleles . Na (Freq >= 5%) = No. of different alleles with a frequency >= 5%. Ne = No. of effective alleles = 1 / (Sum pi^2) I = Shannon's Information Index = -1* Sum (pi * Ln (pi)). No. Private Alleles = No. of alleles unique to a single population. No. LComm Alleles (<=25%) = No. of locally common alleles (Freq. >= 5%) found in 25% or fewer populations. No. LComm Alleles (<=50%) = No. of locally common alleles (Freq. >= 5%) Found in 50% or fewer populations. He = Expected Heterozygosity = 1 - Sum pi^2. uHe = Unbiased Expected Heterozygosity = (2N / (2N-1)) * He.

WM WA WF WZ Population Mean ± SE Mean ± SE Mean ± SE Mean ± SE Na 6.462 ± 0.896 4.154 ± 0.492 4.692 ± 0.683 5.692 ± 0.720 Na Freq. >= 5% 2.923 ± 0.431 3.154 ± 0.390 3.462 ± 0.489 2.538 ± 0.268 Ne 2.354 ± 0.303 2.396 ± 0.306 2.724 ± 0.420 2.056 ± 0.218 I 0.924 ± 0.156 0.915 ± 0.147 0.988 ± 0.175 0.887 ± 0.119 No. Private Alleles 1.385 ± 0.417 0.308 ± 0.133 0.154 ± 0.104 0.846 ± 0.274 No. LComm Alleles (<=25%) 0.000 ± 0.000 0.000 ± 0.000 0.000 ± 0.000 0.000 ± 0.000 No. LComm Alleles (<=50%) 1.231 ± 0.303 1.308 ± 0.328 1.308 ± 0.347 1.231 ± 0.303 He 0.462 ± 0.077 0.477 ± 0.075 0.492 ± 0.083 0.445 ± 0.058 uHe 0.464 ± 0.078 0.485 ± 0.076 0.499 ± 0.084 0.449 ± 0.058

Genetic diversity of Moringa peregrina revealed with microsatellite markers

4 Genetic diversity and conservation biology of Moringa

peregrina populations growing in Sinai, Egypt

Conclusion and perspectives

This PhD thesis is based on a population monitoring (2006-ongoing) of the four remaining populations of the endangered tree Moringa peregrina (Moringaceae) at the Sinai Peninsula, Egypt. Population dynamics of M. peregrina in Sinai is negatively affected with the impacts of the anthropogenic activities (grazing, cutting, collection, etc) and climate change. The latter phenomenon was the target of a first paper. The climate in the region has always been affected by extremes (droughts for several years; five to seven years without considerable rainfall alternating with sudden flash floods that occurred after these drought years). In addition, we identified long-term trends, like an increase in mean annual temperature (+2.3 oC in the last four decades), A second paper focused on the consequences for M. peregrina: we found a poor recruitment status leading to a reduction in population size of ca. 15 % (from 448 to 390 trees within the last 10 years) with lacking recruitment (only 12 saplings are recorded in 2014 after a strong rainfall in 2010 and completely disappeared in 2015). Not only fitness parameters (number of pods and seed set), but as well tree performance parameters (increase of height, diameter; DAG and DBH, crown area, ordering, and vitality) were affected negatively. The remaining populations were fragmented into four wadis only (Agala, Feiran, Meir, and Zaghra). A re-assessment of the threatened status according to the IUCN criteria, resulted in an assignment to the category endangered (EN) instead of the previous assessment (VU, vulnerable) that was done in 1990.

Furthermore, the response to nine variables; climatic (temperature and annual precipitation), environmental or habitats (elevation, inclination, exposition, landform, and GPS coordinates) and anthropogenic (browsing, cutting, and collection) showed that once established, the trees are relatively resistant to any threats due to the environmental change. However, tree performance was significantly better in the wadi beds, where the water supply is better. At the same time, the habitat choice for the tree reflects a trade-off: the wadi beds are also the habitats most threatened by floods and grazing.

It could be shown, and this is the focus of a third paper (not yet published) that M. peregrina forms annual growth rings. In very dry years, no rings may be formed. This trend in addition to the mostly multistemmed growth form of the trees were challenging the age determination, even if usually, due pronounced seasonal changes, M. peregrina produces only one ring per year. This is was proved from the relationship between the tree radius and the annual growth rings that could be used as a reliable marker for age dating of the remaining populations in two different surveys (2007 and 2015). To overcome possible inaccuracies in age estimation, three different methods were used. Methods A (mean of all radii) and B (calculating a virtual single radius) tend to overestimate age for multi-stemmed trees, whereas method C (counting only the thickest stem) may 117

Conclusions

underestimate it. Nevertheless, all three scenarios revealed missing recruitment over prolongated periods of time (mainly after years with severe drought). Based on the the estimated mean age difference from two censuses, method B can be recommended. Based on the number of the established trees, failure years in which the population could not establish new individuals were counted. All used methods of age dating pointed to longer periods of recruitment failure, with 62 years (1839–1901) and the last 32 years (1983–2015) for W. Agala. Similar long periods are indicated for the other wadis: W. Feiran, 28 years (1841–1869) and the last 27 years (1988–2015); W. Meir 26 years (1725–1750) and the last 36 years (1976–2015), although it comprises the largest population in Sinai (242 trees); W. Zaghra 68 years (1659–1727) and the last 30 years (1985–2015). In combination with the anthropogenic impacts and climate change especially in the last four decades, natural fluctuations in recruitment pose a high risk for the remaining populations that are in severe decline. Unless the management and rehabilitation programs, the current remaining populations of M. peregrina will be completely extinct.. Form that reasons, an action plan and listing of M. peregrina in the IUCN Red List was highly recommended.

The genetic diversity of the populations was investigated with 13 SSR markers; results are to be published in a fourth paper. These markers showed a high resolution especially the mean error rate was only 2 % per MLG. The genotype accumulation curve as well demonstrated that 7–8 markers were necessary to discriminate between 100% of the multilocus genotypes. Results point to a rather high degree of genetic variance within individuals or within populations although the populations are suffer from genetic isolation and inbreeding within a population. An analysis of molecular variance (AMOVA) revealed 17% of the genetic variation among populations, 8% among individuals and 75% within individuals or 83 within population if the analysis within the individuals’ analysis suppressed. Significant departures from HWE was detected for eight loci (P < 0.001), probably due to a high degree of inbreeding within a population. Data also showed low gene flow between three of the four populations, an exception occurred for the two wadis which are geographically close to each other (around 1 km distance) in addition few outliers due to the translocation of seeds by Bedouins and researchers from wadi to another. The analysis of genetic structure under different Ks (K between 1 and 7) showed that most pronounced genetic structure was for K=3 with lumped populations from two neighbouring wadis (W. Agala and W. Feiran) together. The three groups seem to be now genetically isolated. They may be remainders of a formerly contiguous population, especially when considering the change towards a drier climate in Northern Africa within the last 6000 years. Six clones revealed some degree of asexual propagation, most likely occurring via broken twigs, which may have rooted after flash floods. It may be an alternative mode of reproduction under harsh conditions.

Based on the abovementioned results the following activities are recommended: (1) Regeneration and conservation programs (preferably overseen by local Bedouins which have an interest to utilize the trees) should be initiated, focusing on facilitated recruitment of young individuals in selected natural habitats. This program should include as well the protection of the old trees to preserve genetic diversity within the trees of the area. Collecting and germinating seeds from the largest and oldest trees in W. Zaghra and W. Meir would be a first measure. 118

Conclusions

(2) Continuing the two censuses (2007 and 2015) already carried out, a long-term monitoring of population structure would be desirable. Together with records on years with extreme weather events (drought or local flash floods), this would enable a population viability analysis (PVA) which is urgently needed to identify the drivers of population dynamics for the tree. (3) We recommend as well to include Moringa peregrina in the IUCN Red List of Threatened Species. Since M. peregrina is endangered due to several threats, mainly anthropogenic impact and climate change, which may act in a synergistic manner, and it was difficult to predict the risk of extinction. (4) To minimize risks of the bad conservation status, an action plan by the Saint Catherine Protectorate Authority would help in mitigating the adverse impacts of these threats. (5) Conserving the genetic resources for the remaining populations of M. peregrina in Sinai; giving priority to the most diverse population situated in W. Zaghra. Conservation could include the deposition of seeds in seed banks, in situ, and ex-situ conservation programs. (6) Perform paternity analysis, comparing offspring raised from collected seeds with mother trees (the 13 SSR markers established in this study would easily allow that) to determine the mating systems of the tree, especially if the species is obligatory or facultatively allogamous. (7) The degree of outcrossing could as well be estimated with pollination experiments. In addition, artificial pollination may help to overcome possible inbreeding effects. To facilitate acceptance of conservation programs by the local Bedouins, offspring of trees with high pod weight and/or oil content could be planted around settlements, which may be a first step to select genotypes for direct cultivation.

119

Declarations

120 120

DeclarationsConclusions

Eigenständigkeitserklärung (Diese Erklärung ist zu unterschreiben und in die Dissertationen einzubinden.)

Hiermit erkläre ich, dass diese Arbeit bisher von mir weder an der Mathematisch-Naturwissenschaftlichen Fakultät der Ernst-Moritz-Arndt-Universität Greifswald noch einer anderen wissenschaftlichen Einrichtung zum Zwecke der Promotion eingereicht wurde. Ferner erkläre ich, dass ich diese Arbeit selbstständig verfasst und keine anderen als die darin angegebenen Hilfsmittel und Hilfen benutzt und keine Textabschnitte eines Dritten ohne Kennzeichnung übernommen habe.

I hereby declare that I did not previously submit this work at the Ernst-Moritz-Arndt University Greifswald or any other scientific institution as a doctoral dissertation. I further declare that I completed this work independently and did not use any aid or means except the ones specified. All sources and means are specified in detail in the respective publications and the declaration of my contribution to the collaborative works.

Unterschrift des Promovenden

Greifswald, 14.03.2018

121

ConclusionsDeclarations

Declaration of my contributions to the publications

This dissertation includes a total of 2 peer-reviewed publications and 3 manuscripts with contributions from Mohamed A. Dadamouny (thesis author), Prof. Martin Schnittler (Ph.D. Supervisor) and the following colleagues; Dr. Martin Unterseher and Dr. Peter König in the second chapter and Mrs. Anja Klahr and Mrs. Franziska Eichhorn helped in the lab work for results presented in the fifth chapter. The following paragraphs list all author contributions in detail.

Chapter 3.1. Dadamouny MA and Schnittler M (2016) Trends of climate with a rapid change in Sinai, Egypt. Journal of water and climate change 7(2):393–414. doi: 10.2166/wcc.2015.215. MD collected data from the field stations in Sinai and from the Egyptian Meteorological Authority (EMA) in Cairo. MD also arranged all and evaluated these with different statistical programs. MD wrote the initial manuscript and figures as corresponding author. MS suggested some analysis methods and helped with data evaluation. Both authors edited and revised the final version of the paper.

Chapter 3.2. Dadamouny MA, Unterseher M, König P, Schnittler M (2016) Population performance of Moringa peregrina (Forssk.) Fiori (Moringaceae) at Sinai Peninsula, Egypt in the last decades: consequences for its conservation. J for Nature Conservation 34:65–74. doi: 10.1016/j.jnc.2016.08. MA and MS designed the idea and the methodologies used for this publication. MD (the corresponding author) carried out field work and collected the data. MD and MS did the data evaluation and identified the necessary analysis methods. MD did the initial analysis of the data and wrote the initial manuscript. MS and MU contributed to the data treatment and multivariate analysis using R program. MD and PK collected the distribution data and MD did the IUCN assessment of the threatened species. MD and MS edited and revised the paper.

Chapter 3.3. Dadamouny MA and Schnittler M. Demographic and size structure status of the threatened populations of Moringa peregrina trees at Sinai Peninsula, Egypt. (Manuscript). MD collected all tree and environmental data in two censuses 2007 and 2015, collected the wood samples and analyzed the data. He wrote the initial manuscript. MS suggested the study and proposed the three methods used for age determination of trees. Both authors interpreted the data, edited and revised the paper.

Chapter 3.4. Mohamed Dadamouny, Anja Klahr, Franziska Eichhorn, and Martin Schnittler. Genetic diversity of the endangered species Moringa peregrina (Moringaceae) in Sinai, Egypt revealed with microsatellite markers. (Manuscript). MD and MS designed the study and conceptualized the project. MS suggested methods for sampling. MD carried out surveys and a comparative analysis of the methods used for genotyping, draw the maps. MD carried out fieldwork and collection of samples for DNA extractions. MD extracted DNA from the samples collected from Sinai using different methods and AK helped in optimizing the methods of the DNA extraction. MD establishing the assay and carried out the lab work (DNA was measured and diluted, PCRs 122

Declarations Conclusions for ITS and PCRs for cross-amplification, testing primers, and optimizing the PCR conditions, sequencing the PCR products required to check the SSR motifs, and multiplex PCRs). FE gave advises for DNA extraction and data evaluation on the GeneMapper. MD did scoring and Tandem binning, which was revised by MS and FE. MS, MA, and EF contributed to data evaluation. MS analysed the data for clones. MD calculated all population genetic parameters and wrote the initial manuscript which was revised by MS and MD.

Mohamed A. El Dadamouny Sheteiwy

Date: 14……………….03.2018…………. (Author of the thesis)

Obige Angaben werden bestätigt:

Prof. Dr. Martin Schnittler Date: 21…………………………….08.2018 (Wissenschaftlicher Betreuer)

123

ConclusionsDeclarations

Erklärung zur Abgabe einer elektronischen Kopie der Dissertation

Mathematisch-Naturwissenschaftliche Fakultät Einverständniserklärung nach § 4 Abs. 1 Nr. c Promotionsordnung

Hiermit erkläre ich, dass von der Arbeit eine elektronische Kopie gefertigt und gespeichert werden darf, um unter Beachtung der datenschutzrechtlichen Vorschriften eine elektronische Überprüfung der Einhaltung der wissenschaftlichen Standards zu ermöglichen.

Datum: 14.03.2018

Mohamed A. El Dadamouny A. Sheteiwy

124

Acknowledgements

Acknowledgements

First of all, I would like to thank our God, without his grace and his care I could not achieve the present study. Also during the course of this thesis, many people have supported me. It is a very pleasant and honour moment for me to express my sincere gratitude to all of them.

I would like to express my deep gratitude and appreciation to my supervisor Prof. Dr. Martin Schnittler (director of Institute of Botany and Landscape Ecology, University of Ernst-Moritz-Arndt Greifswald). I could not put into words what he did for me since my coming to Germany. I appreciate his guidance, support, motivation, and encouragement throughout the course of this Ph.D. project and initiating my research work. Mine acknowledge also extends to Dr. Martin Unterseher and Dr. Peter König, Institute of Botany and Landscape Ecology, the University of Ernst-Moritz-Arndt Greifswald for their help in my second paper on the conservation biology of my studied species and updating its status in the IUCN Red list of the threatened species.

My grateful also to Mrs. Anja Klahr for guiding in the lab work and successful establishment of my experiments and Mrs. Handt for her help in germination and cultivation of Moringa collected from Sinai to be established at the botanical garden, Greifswald as long as my experiments and collecting the required samples for DNA extraction. Many thanks for all staff in our working group for their help and encouragement; Dr. Pascal Eusemann, Dr. Thomas Hoppe, Dr. Nikki Dagamac, and Dr. Abu Bakar Siddique. Also, the secretary of the Institute Mrs. Angelika Elsner for her help and encouragements. I would like to thank Mr. David Würth and Mrs. Franziska Eichhorn, my colleagues in the working group, for sharing their experiences in microsatellite analysis.

I would like to thank Prof. Dr. Werner Weitschies, Dean of Faculty of Mathematics and Science and all staff of Ernst-Moritz-Arndt-Universität Greifswald for these excellent research-oriented facilities and the highly academic Institutes of research and prestigious graduate education. Thanks to Mrs. Brandenburg in the international office and Mrs. Imme Burkart-Jürgens, in the welcome centre for arranging the international meetings and events that helped me to build my network with the colleagues all over the world.

My appreciation to the team of German Academic Exchange Service DAAD (in Bonn and in Cairo) for their support and for this great opportunity to pursue my Ph.D. study in Germany in one of the most promising research areas in population biology within the frame of German-Egyptian Research Long-term Scholarship (GERLS) Program. Great thanks to Dr. Thomas Prahl for his supportive meetings. I appreciate all supportive contact persons; Mrs. Margret Leopold, Mrs. Saoussen Louati, Mrs. Andrea Gerecke, Mrs. Rebecca Gottschalk, and Mrs. Miriam Mbae for their cooperation and support during my scholarship. Appreciation extends also to Ministry of Higher Education, Egypt. I would like to thanks, all staff and professors who recommended and supported my Ph.D. project; Prof. Dr. Abdel Raouf Moustafa, Suez Canal University, Faculty of Science, Ismailia, Egypt.

125