Cedrus is a genus distributed across three disjunct areas: the Atlas Mountains of North Africa; mountainous parts of the eastern Mediterranean basin, in southeast Turkey, northwestern most Syria, Lebanon and Cyprus; and in the Hindu Kush, Karakoram and western Himalayan mountain ranges. (Farjon 2017). C. deodara is well defined, isolated in the western Himalayas, and broadly accepted. The taxonomic position(s) of the populations in the Mediterranean basin is controversial. They have traditionally been treated as three distinct species (e.g. Debreczy & Rácz 2011; Farjon 1990; Davis 1965), often as two (e.g. Farjon 2017; Tutin et al. 1964), and more recently several molecular and phylogenetic studies have even suggested that the Mediterranean Cedars represent a single species (C. libani; the taxonomic implication being that the other ‘traditional’ species should be treated at infraspecific rank e.g. C. libani subsp. atlantica) (Jasińska et al. 2013).
Differentiation of the Mediterranean populations based on morphology alone is particularly difficult and more or less impossible between some cultivated trees of C. atlantica and C. libani (Farjon 2017). Nevertheless, these two traditionally recognised species occur in vastly disjunct areas and each has its own cultural history, both in their native ranges and in cultivation, and so we have opted to continue to recognise them as distinct. C. brevifolia continues to be the source of much debate and its probable overlap with C. libani subsp. stenocoma is discussed in the taxonomic notes of both taxa. Until the status of these two entities is more firmly resolved we have again opted to continue to recognise them both, following the recent ‘four species’ treatment of Debreczy & Rácz (2011). To maintain such a view into the third decade of the 21st century might charitably be described as conservative, however, we believe that recognising these entities in this way better serves our predominantly horticultural audience.
The three species of the Mediterranean basin are, in their wild state, generally monopodial in forest settings, but in areas of their native ranges with a long history of disturbance – and especially in cultivation – they are inclined to be multistemmed, often with several first order branches becoming co-dominant with the main stem, especially in C. libani, and with massive lateral branches from close to the base developing in open-grown specimens. C. deodara exhibits a greater degree of apical dominance than the other species which may account for this reaching the tallest heights among the cedar species.
Evergreen trees to 65 m tall, to 4.5 m dbh. The bark is thin, grey-pink-brown, later darkening and cracking into small plates, eventually becoming thick on very old trees. Young plants are conical, with regular branching. The branchlets have marked dimorphism: leading shoots are long and slender (‘long shoots’), lateral shoots short and stunted (‘short shoots’); vegetative buds are not resinous or only slightly so. The leaves are needle-like, dark green to glaucous-blue, rigid, spirally arranged, slightly narrow at the base, the apex acute. The male strobili are erect, solitary, catkin-like, terminating short shoots, and thus subtended by a whorl of leaves. The female cones are erect, also positioned on short shoots, less abundant than the strobili, but both are usually located on the outer branches around the crown; the female cones mature in the second year, glaucous-green turning dull brown, disintegrating on the tree. The seed scales are broad, somewhat papery, subtended by insignificant bract scales, spirally arranged around the central rachis; at maturity the scales fall to leave the rachis. Seeds are borne two per scale, partially enclosed in a membranous cup, which extends to form a wing. There are few morphological characters to quickly and reliably separate the species, though C. deodara is relatively distinct from the others, and habit characteristics are useful with cultivated material. (Farjon 1990; Debreczy & Rácz 2011).
Only the most committed, unflinching atheist could gaze upon a natural stand of mature Cedars and fail to understand why they are so often called the trees of God. This extraordinary phraseology may owe its origins to the small, remnant stand of Cedrus libani at Bsharré in Lebanon, which as a specific stand is sometimes known as ‘the Cedars of God’, though whether it was so-Christened by visitors, or the inhabitants of those lands in the past, is unclear. Similarly, in the western Himalaya, the specific epithet deodara was derived from the Sanskrit word devadāru which means ‘wood of the gods’.
At least three of the Cedars, Atlas, Lebanon, and Deodar, all enjoy the distinction – rare among the more iconic conifers grown in temperate gardens – of attaining in cultivation an appearance that at least comes close to matching what one can expect from them in the wild. No cultivated trees of any of these three species can match their wild counterparts in size, but mature Cedars lend an air of antiquity to our landscapes that no other conifer can, except perhaps an ancient Yew.
Despite their restricted natural distributions the Cedars have established themselves as a truly cosmopolitan genus in horticulture. Combining heat tolerance and at least moderate cold-hardiness with their stately appearance, they have become an extremely important group of ornamental trees and may be found in parks, gardens, and a variety of public-green-spaces throughout the mediterranean and temperate zones of the world (Dirr 2009, Pijut 2000). They appear to be limited by only the most severe winter cold, which can prohibit the establishment of young trees in areas with a truly continental climate, for example in parts of North America, Scandinavia, and at northerly latitudes in Asia. Multiple attempts to grow then at the Dominion Arboretum in Ottawa through the mid-late 20th century all ultimately met with failure (Buckley 1980). They are widely grown almost everywhere else, except in the tropics and the truly arid regions. All four species have been heavily exploited in their native ranges for their timber, which is strong and durable, and there is evidence in the engravings on the Palermo Stone that C. libani has been exploited for its timber for nearly 5,000 years.
Legal and illegal logging continues to affect all species and in recent years there has been much interest in the genus in Europe, as state and commercial foresters seek to mitigate against the projected effects of climate change and introduce new species to industrial-scale plantations for timber production. Early small-scale plots of Cedrus have shown promise in many areas of Europe, including in the southern part of the United Kingdom (Savill & Wilson 2015). In ornamental situations Cedars perform best in areas with moderate summer heat, on deep, fertile, moist but freely draining, loamy soils. They will also tolerate clay soils so long as they are not waterlogged, and they can grow well on shallow soils over limestone, chalk, etc., for example at Goodwood, Sussex, and at Pampisford Hall, Cambridgeshire, both in the United Kingdom (pers. obs.). They typically require a sunny, open position, and particularly when young will not tolerate anything more than light shade. (Rushforth 1987a). In England, C. libani generally performs best in the southern half of the country but is perfectly capable of growing into a stately tree even in the north of Scotland if properly sited. C. atlantica would seem to be the hardier of the two, or at least the more capable of succeeding in harsher conditions, but like all Cedars it rarely thrives in poorly drained soils (Savill & Wilson 2015).
C. atlantica is also far more tolerant of air pollution than C. libani, a trait first noticed in cultivation in western Europe as the landscapes around major cities gradually became industrialised through the latter half of the 19th century, therefore it is this species that is most often advocated for use in urban settings (Pijut 2000; Dirr 2009). C. deodara too is tolerant of air pollution, as demonstrated by the great density of many fine trees, some of considerable age, that still grow across the southeast of England including the London area. All four species are cultivated in North America, but C. libani is scare here compared with Europe, while C. atlantica and C. deodara are relatively common. C. atlantica is valued in North America for its tolerance of the intense summer heat of the southern United States, though in northern areas it needs to be grown in a sheltered position with protection from cold winter winds (Dirr 2009).
While the Cedar of Lebanon and the Atlas Cedar are quite similar in their general appearance and even their gross morphology – and indeed faced with a mature example even an expert can be hard-pressed sometimes to be absolutely certain as to which is which (Farjon 2017) – the Deodar and Cyprus Cedars are quite different. Lebanon and Atlas become massive trees, often with co-dominant leaders and gigantic branches striking out from close to the bole, with an extensive, spreading canopy. The horizontal branching structure creates the distinctive crown, which is often divided into a series of sub-crowns, each flat-topped, sometimes separated from one another by the architecture of the tree which will, invariably, have been crudely pruned by at least two or three severe storms in its lifetime. These are the ‘classic’ Cedars so beloved in many of our historic landscapes.
The Cypriot is the rarest in cultivation. In many ways can be looked upon as a miniature Cedar of Lebanon, at least in terms of its functionality as an ornamental species in gardens, and whether or not it deserves recognition at species rank has been debated for some time (Farjon 2017). It can almost always be distinguished by its smaller stature and by its consistently shorter needles across branches of varying ages. To help us further, foliage on its young shoots is often tinted bronze (pers. obs.). C. deodara, while instantly recognisable as a Cedar, is quite different. It is set apart by the distinctly pendulous leading shoot and lateral shoot-tips, the longer needles, and the apical dominance that very often gives Deodars a different profile as a mature tree – though still very stately – to the other species. Hailing from the Western Himalaya where it has a large east-west and altitudinal distribution, the Deodar is very adaptable provided the correct provenance is selected. It is a common ornamental in both Scotland and Uruguay, for example (Mitchell 1996; Knap 2003).
Propagation is by seed, except the cultivars that must all be reproduced by grafting. Viable seed is set in cultivation in warmer areas, and this can happen in the United Kingdom and Ireland given a warm summer, such as the summer of 2018. However, seed of cultivated trees should be viewed with suspicion as it is possible that hybridisation could have occurred (see for example C. atlantica × deodara and within that species the cultivar ‘Ibrido’). Indeed, hybridisation between Atlas and Lebanon Cedars has been known since at least 1930 (Jacobson 1996) but these generate less excitement, possibly because the parents can be so hard to distinguish in the first place.
Glaucous forms frequently occur in batches raised from seed, and it is worth pausing for a moment to dispel a myth that only Atlas Cedar yields glaucous forms. Certainly, it is the glaucous forms of this species, C. atlantica Glauca Group, that have become such very popular trees in cultivation, but there are Blue Atlas Cedars and Blue Atlas Cedars, so care should be taken to ensure that the strength of colouration is as desired when purchasing plants. In addition there are named glaucous forms of Deodar, and indeed of Cedar of Lebanon, C. libani ‘Glauca’, a very rare form indeed that originated in 1855 and was for a long time known only from the remarkable tree at Pampisford Hall, Cambridge, UK (Mitchell 1972). A very great many cultivars have been selected from the Cedars and we are indebted to Aris Auders and Derek Spicer (Auders & Spicer 2012) for their scholarly work, which shines a light on the often vexatious characterisation of conifer cultivars. Their encylopaedia has been referenced heavily in the cultivar accounts under the various species.
Returning to the types, however, it must be noted that all too often nurseries, even the ones that ought to know better, sell Cedars too large. There can be few more depressing sights than a young Cedar that has been grown too long in too small a pot (and not even an air-pot at that), mercilessly staked to force a grafted plant to form a leader, and then planted and expected to turn into a tree of unrivalled beauty. This ‘deplorable practice’ as Alan Mitchell called it (Mitchell 1996) can and must be stopped. It is the dendrological equivalent of hurling paint at the wall directly from the can and expecting wonderful results. Cedars should ideally be planted when relatively young, 3–6 years is ideal, but no older, unless they have been expertly grown and have as healthy and well-structured a root system as they do upper growth. The first lateral branches should already be well developed, and the root system should be healthy – unless it is a cultivar that has been grafted it will not need staking. It takes pot-bound Cedars a long time to grow out of their tortuous nursery experience and the duped customer will spend many years wondering why his Cedar is looking so uninspired.
Cedars are as palatable as the next tree to herbivores so need to be adequately protected, and as young trees they are particularly sensitive to competition from grass, so if they are being planted in a grassy area the grass must be eradicated close to the young tree and prevented from re-growing by the use of mats and/or mulch. For many years Cedars tended to be relatively free from significant pests and diseases, but more recently severe shoot blight and defoliation has begun to appear, particularly on Atlas Cedar. This is caused by the fungus Sirococcus. There are several species of Sirococcus that affect different conifers, but the one most damaging to Cedars is S. tsugae, which is believed to have arrived in Europe from the Pacific northwest of North America in the early 2000s. One of the earliest confirmed cases in England was on an Atlas Cedar avenue at Hergest Croft in Herefordshire, and many of the trees that once made up this avenue have since been removed (pers. obs. 2010–2019).
After its discovery in England S. tsugae has since been found at further locations in England, Wales, Scotland, and Northern Ireland, and it is present in continental Europe, too (forestresearch.gov.uk). Currently there is no officially sanctioned treatment for the fungus, but private individuals have found that treating young trees with a general-purpose fungicide as they come into growth, and again later in the season, can help (anon., pers. comms. 2018–2019). The fungus is most damaging on C. atlantica, damaging but less so on C. libani and C. brevifolia, while C. deodara seems to show a higher degree of tolerance. S. tsugae also affects several of the North American Tsuga species, T. heterophylla, T. canadensis and particularly T. mertensiana.
A great challenge posed by Cedars is keeping them structurally sound as they grow old. With their enormous mass, complex architecture, and propensity to be damaged by gales or by falls of wet, heavy snow in maritime climates, Cedars in parks and gardens and other spaces regularly accessible to the general public can pose a management nightmare. The same may be said for Cedars growing in close proximity to buildings, roads, or other important structures and thoroughfares, whether or not there is public access, and in all such situations it is imperative that land owners and managers have such trees regularly inspected by a professional arboriculturist, that records of these inspections are kept, and that any recommended management actions are undertaken as necessary. Cedars are sensitive to soil compaction, and on soils that are disposed to compaction, care must be taken to prevent this from happening, or to alleviate compaction that has already occurred.
Finally, we must turn to the vexing question of how to assuredly distinguish the different species from one another. This is challenging enough in the United Kingdom, but one must be doubly careful in areas with consistently warm summers, where Cedars will regularly set fertile seed which raises the possibility of hybridisation having occurred. Horticultural students are sometimes advised to use alliteration to identify Cedars: ‘ascending Atlas, level Lebanon, drooping Deodar’ (i.e. Jacobson 1996), referring to the direction of growth of the youngest shoots. This might be sufficient to get through one’s plant identification exams, but later on it becomes clear that it isn’t so simple. Trees of Atlas and even Lebanon Cedars producing new extension growth during a warm, dry spring, can often have distinctly drooping new branchlets that could throw an inexperienced observer off the scent, and this problem is exacerbated in regions with warmer summers. Furthermore, in their natural habitat, Deodar Cedars can have branches with more or less flat tips (Ryan 1999) but this is almost unheard of in cultivation.
C. deodara is the most immediately distinctive of the four, with its (usually) pendulous shoot-tips and the longest needles of any Cedar. The conveniently named C. brevifolia can also be quickly distinguished by virtue of its consistently short needles – once seen never forgotten – and often by the bronze tint to its long shoots. It is C. libani and C. atlantica that are most alike and which can be the most challenging. In the United Kingdom, we know libani was cultivated long before atlantica: The accepted dates of introduction are 1638 and 1840 respectively, so if one is lucky enough to be dealing with trees with a known planting date prior to 1840, then one can confidently say Lebanon. Trees planted after this date, though, which most in the United Kingdom are, can be very difficult and close examination is required. Many remarkable records for both these species from the website monumentaltrees.com have not been mentioned in the species accounts because in most cases we cannot be certain of the identification, even when records are illustrated there often remains ‘reasonable doubt’. This uncertainty is a pity as that website is a remarkable resource, and many of the trees recorded and illustrated on it are simply staggering, regardless of their identity.
Several authors have offered various tips for tackling this problem of confidently separating Atlas and Lebanon Cedars, including:
‘In the Atlas cedar, the shoots are always downy, and more so than those of the Lebanon species. Its cones, too, do not taper above the middle so much’ (Bean 1981a);
‘The Lebanon Cedar has longer needles…usually sharper, flushing earlier in spring and contrasting their bright new green colour with the dark old needles. Cones larger…and fewer; bark darker; twigs far less hairy. Branching usually strongly tabular whether narrow or broad [in crown]’ (Jacobson 1996);
‘The distinctions between [C. atlantica] and C. libani are not absolute, but include habit; angle of branch ends; length of leaf; translucent spine on leaf; smaller cone and broad cavity at the top of the cone, and grey rather than red-brown bark. The one distinction that is quite invalid is the colour of the leaf…With several features taken together probably all trees can be distinguished and at present it seems that the translucent spine on the leaf-tip is the most constant, but time of shedding pollen might be valuable’ (Mitchell 1972);
And more recently the translucent spine on the leaf-tip has been described as ‘reddish 0.2 mm’ in C. libani, and ‘yellowish 0.3–0.5 mm’ in C. atlantica (Poland & Clement 2009).
The key to the species given here is adapted from Farjon (1990).
Leading shoot and branch tips strongly drooping, nearly pendulous. Trees usually single-stemmed, strongly upright. Leaves 2–6 cm long. Male strobili usually over 5 cm long.
Leading shoot and branch tips erect, spreading, or depressed (not drooping). Trees with one or more major stems or with major branches from low down. Leaves equal to or less than 3.5 cm long. Male strobili shorter than 5 cm.
Male strobili 4–5 cm long. Leaves (1–)2-3.5 cm long, on short shoots in false-whorls of 20–35. Trees often multi-stemmed or with massive lateral branches spreading horizontally, occasionally with long clean boles.
Male strobili 3–4 cm long. Leaves usually not more than 2.5 cm long.
Leaves 0.5–1.3 cm long (occasionally longer on long shoots) in false whorls of 15–20. Branches spreading more or less horizontally.
Leaves 1–2.5(–3) cm long in false whorls of 20–45. Usually single-stemmed with horiztonal branching.