Short Shoot Growth and Reproduction Response to Light Conditions Vary with Order Branching in the Proximal Part of C. atlantica Crown

0is study compared the effects of shading in individual branch orders 2 and 3 on the needle survival, growth, and reproduction of five categories of short shoots of the proximal part of wild Atlas cedar (Cedrus atlantica (Endl.) G. Manetti ex Carrière). 0e sun exposure did not affect the number of short shoots in the two branch orders, whereas light compared to shade only stimulates the unbranched short shoot elongation of the branch order 3. 0e impact of shade exposure compared to sun on the loss of needles depends on the order of branching; it is weak to order 2 and increases to higher order. 0is effect in the branch order 3 is achieved by a significant decrease of the fallen leaf number in the unbranched short shoot SSnr and the short shoot SS/T worn by Twigs while in the branch order 2 only short shoot SS (nr + r) loses significantly few needles. In terms of short shoot extension and needle loss, the SS/Tof the branch order 3 behaves in the same way as the SS (nr + r) of the branch order 2.0e shadow compared to sunlit only decreases significantly the production of pollen strobili of the branch order 2. Close relationships between short shoot extension, leaf life span, and pollen strobili production of axillary products in the proximal part of C. atlantica crown were found.


Introduction
Beyond a broad range of species, the relationship between location and the fate of axillary buds controls the branching process which is known to be a more complex event [1]. is process has been extensively addressed in many species of deciduous and conifers and a decrease in the length of the laterals has been observed from the top of the bearing shoot [1][2][3][4][5][6]. In part of these species, some shoots depend on their location and age, and the numbers of laterals and branching characteristics were found to decrease with plant development and ageing in many woody species, including C. atlantica [7][8][9][10]. Similarly, on six apple cultivars, Costes and Guédon [1] reported that branching is organized along the annual shoot according to particular lateral type locations. Moreover, Millet [11] and Vennetier et al. [12] have found that, for young to adult conifers, the annual shoot length also decreased from the hierarchical order 1 to the following orders 2 and 3. Colombo and Templeton [13] have revealed that the annual shoot length was strongly correlated with branch position in the crown, thus with branch age in pines and spruce. In the same way to the latter authors, Buissart et al. [14] demonstrated that the annual shoot length and the total number of branches and whorled branches were highly variable and decreased with age and with increasing branching order for Picea mariana, Pinus banksiana, and Pinus strobus. In contrast, Remphrey and Davidson [15] announced that generally, older branches tend to grow slowly whereas lengths of current shoots of major order 1 branches were found to decrease with increasing depth in the crown [16].
Several studies have been done on the effect of light on growth [17][18][19]. Aussenac [17] has shown that the decrease in light produces a reduction in the palissadic tissue of the needles of four conifers. In a study of Suzuki [18], it was found that Cleyera japonica annual shoot length and light environment were positively correlated within the same order branches. Recently, on Picea glehnii, Chen and Sumida [19] showed that the production of current-year shoots on a primary branch increased with increasing light whereas secondary branches at the distal part of a primary branch produced more current-year shoots and exhibited a lower death than proximal branches. While they note that, for the primary and secondary branches, the mean shoot lengths in the lower crown were not significantly different between sunlit and shaded trees and they conclude that local light intensity alone cannot fully explain the growth and survival of primary branch in contrast to secondary branch. On the other hand, it has been shown that the age of the oldest leaves (longevity) on a branch of conifer leaves varies with species and habitat conditions (elevation, latitude, fertilization, and irrigation) and such intravariation is largely interpreted as environmentally determined phenotypic acclimation [20]. Furthermore, it was reported that different-aged needles of conifers affect the net photosynthesis and significant differences in mobile sugar concentrations in these needles [21] and the correlation between an increase in tree height or age and a decrease in CO 2 assimilation rates were found [22].
Cedar growth has been found to be relatively long when compared to other resinous such as fir [23]. e branching process in adult and young Cedrus atlantica (Endl.) G.Manetti ex Carrière tree has been studied by counting the number, length, and position of short and long shoots [24,25]. Other studies have been made on growth dynamics and morphology of annual shoots according to their architectural position in young C. atlantica [8] and have found that annual final shoot length, duration of extension, and maximum extension rate decreased with increasing branching order. e inter-and intraspecific variation of the polycyclism was found in Cedrus, to genetically discriminate between Cedrus species, and is considered a way of adaptation to adverse climatic conditions [26]. Like for Eurya japonica [27], Sabatier and Barthélémy [8] have found that annual shoot length decreased with increasing branching order for C. atlantica (Endl.) and have reported that the types and numbers of sylleptic axillary shoots produced are linked with both final length and extension rate of the parent shoot in young C. atlantica. Similarly, Courbet et al. [28] have constructed a model to predict branch vertical position which can be used as inputs for branch diameter and mortality models for Atlas cedar. is model shows that the vertical location of branches in the crown partly governs their further growth and mortality from which depend (i) the stem growth and form and (ii) the quality of lumber and veneer, including wood knots.
Many other studies have been made on the cedar response to environment and have furthermore shown that (1) the better light and water availability induce higher growth rates, wider tree rings, and lower leaf area index in Lebanon Cedar [29], (2) the low precipitation during summer limits tree growth on Lebanon Cedar [30], (3) the adverse climatic conditions and soil fertility affect the polycyclism and annual shoot length in Cedrus [26], (4) the low nutrient availability reduces the growth unit size, life duration of branch, and shoot number and modifies the axillary product category and their distribution in C. atlantica [24], (5) the high water status and low global radiation increase the duration of the daily growth period in C. atlantica [23], (6) the dry set decreases height growth for C. atlantica more than C. Brevifolia and vice versa in the well-watered set [31] proving that C. atlantica is the last drought-tolerant compared to C. brevifolia and C. libani. Recently, Bell et al. [32] found that variability of pollen grain size and production within individual samples of Atlas cedar may result from the irregular development of pollen without any evidence for climatic and nutrient influences and concluded that it is influenced by a number of complex factors. Consequently, they underscore the need for further investigation of the complex controls on pollen and strobili development because it would not be possible to use this for the moment as a proxy for environmental reconstitution and cedar conservation.
All these studies show that branching processes and needle death of the genus Cedrus like other species depend on plant development, age, location, and environment but furthermore, might be conditioned by the neighboring branch order in the proximal part of the tree. Moreover, as said above, the effects of conditions on growth and survival appear to be studied extensively [33][34][35]. Besides this, more evaluations of the individual tree and stand leaf formation have been built for several resinous species, but until now no studies have been conducted on the link of branch growth with the fallen leaves and the sexuality of the short shoots in Cedar forest under light and shade exposures. Likewise, no one has attempted to estimate the lifespan of the leaves in the proximal part of the crown of adult trees of C. atlantica and its relationship to the extension of short shoots, the order of branches, and the number of strobili (male cone) per branch although the knowledge of these relationships is fundamental for the study of basic exchange processes in the canopy and the conservation of cedar forests in Morocco. e aim of this study is to understand how light availability affects needle survival, short shoot growth, and their sexuality within an adult tree crown of Cedrus. . is study was conducted on an isolated adult tree of Atlas cedar (diameter at breast height (DBH) � 1.75 m, tree height � 15.08 m) at the edge of the forest but gradually exposed to the sun from morning to evening on these three sides while its fourth side remains in the shade all day from spring to autumn (Table 1). At noon, the light intensity which is measured at the luxmeter at each branch is reported in Table 1. On this tree, twenty-four branches have been sampled at the rate of six per condition of illumination and types of branches: sunlit order 2 branches (L2), shaded order 2 branches (D2), sunlit order 3 branches (L3), and shaded order 3 branches (D3) ( Table 1).

Analysis.
Cedar growth was determined by measuring the length at double decimetre (±1 mm) and counting (i) the number of each type of short shoots carried by the parent branch of orders 2 and 3: unbranched short shoot SSnr carried by parent branch, branched short shoot SSr carried by parent branch, sum of branched and unbranched short shoot SS (nr + r) carried by parent branch, short shoot SS/T carried by twigs of parent branch, short shoot SS/T/LS carried by the twigs of the long shoots of the parent branch, and short shoot SS/LS carried by the long shoots of the parent branch, (ii) the number of existing needles and fallen leaves of the branches L2, D2, L3, and D3 counted under the magnifying glass. e origin of the different types of short shoots mentioned above has been determined as described in Table 2 according to Sabatier and Barthelemy [37] and Sabatier and Barthelemy [26], and (iii) the number of sexual short shoots carrying male strobili and the number of vegetative ones of five types of short shoots of the four experimental branches were counted. e frequency of sexual short shoots was determined by the following formula: frequency � no.of short shoots carrying pollen strobili no.of total shortshoots × 100. (1)

Statistical Analysis.
All the parameters discussed above have been subjected to the analysis of variance (ANOVA), and the means were compared by Fisher's least significant difference (LSD) post hoc test at P < 0.05. All data were analysed statistically using IBM SPSS Statistics software (version 20.0, 2011).

Results
e total and mean number of short shoots carried by the order 2 and order 3 branches showed no significant differences between the branches of orders 2 and 3 neither in light (L2 and L3) nor in shadow (D2 and D3) and between the sunlit branches (L2 and L3) and the shaded branches (D2 and D3) of the same order (Figures 1(a)and 1(b)). On the other hand, the mean length of short shoot SSnr and SS (nr + r) on the order 3 branches is significantly less important on the shaded branches than on the sunlit ones. But no significant difference has been showed on the order 2 branches (Figure 1(c)). e total number of needles and fallen leaves shows no significant difference between shaded and sunlit branches, neither on order 2 nor on order 3 branches (Figure 2(a)). But, when we compare between the effect of light and darkness on the fallen leaf mean number of each short shoot type, we find that short shoot of the order 3 branches lost significantly more needles than the order 2 branches. In the order 3 branches, the needle loss concerns the short shoot SSnr, SS (nr + r), and SS/T, whereas in the order 2 branches the needle loss concerns only the short shoot SS (nr + r) which lost significantly more needles (Figure 2(b)). e analysis of the mean frequency of the sexual short shoot over four experimental branches has revealed that the production of pollen strobili per branch was about five times more in the sunlit order 2 branches (L2) as compared to shaded order 2 branches (D2) (Figure 3). Similarly, two times more production was observed in the sunlit order 3 branches (L3) than shaded order 3 branches (D3). Elevated mean frequency of the sexual short shoot was observed in L2 and L3 branches, with an average number of 5.588% and 1.62% of the total number of short shoots per branch, respectively ( Figure 3). e ANOVA of the mean values of the sexual short shoot frequency per branch has revealed a significant effect of shade (P � 0.032) in the order 2 branches whereas we found no significant effect in the order 3 branches (P � 0.458).
ere is no significant difference between the two orders of branches neither in the shade (P � 0.919) nor in the light (P � 0.093) ( Figure 3). However, interactions of order × light availability were only significant when comparing branches L2 and D3 (P � 0.027) but not branches (L3) and (D2) (P � 0.518). e sunlit order 2 and 3 branches carry five types of sexual short shoot with male strobili at the frequency of 2.15 to 11.39% (Table 3). Both shaded order 2 branches (D2) and particularly shaded order 3 branches (D3) reduced drastically the mean frequency of male short shoots of each type of axillary product as compared to the sunlit order 2 and 3 branches (L2 and L3) ( Table 3). Of all the short shoots carried by the shaded branches of order 3, only those not branched could form a single male strobili, while in the shaded branches of order 2, only two types of short shoots had not formed male strobili: the branched short shoots SSr and the short shoot SS/T/LS carried by the twigs of the long shoots of the parent branch (Table 3). ere was no significant treatment effect (P < 0.05) with respect to the mean frequency of sexual short shoot production per type of short shoot (Table 4).

Discussion
is study shows that neither the order nor the light affects the production of short shoots and that the young branches of order 3 behave in the same way as those of lower order visa-vis of the light availability. Buissart et al. [14] for many conifers established that the total number of branches and whorled branches decreased with increasing branching order. However, Sabatier and Barthélémy [8] have showed that sylleptic axillary shoot occurred only on annual shoots of the branch orders 1 and 2 when the extension rate was at its highest but did not occur in branch order 3. Such results are in congruence with this study and suggest that, in the same light condition, the two studied order branches have reached their full extension rate and thus produced similarly the same number of short shoots in both light and shade  Figure 1: Variation of the total number of short shoots: (a) the mean number of short shoot (b) and the mean length of short shoot (c) among L2, D2, L3, and D3 branches exposed to different light intensities as described in Table 1. Different letters indicate only significant differences among branches (P < 0.05, n � 6).  [38] has found null and strong effect of shade on the production of long shoot in Betula pubescens depending on the year, suggesting that Atlas cedar is less sensitive to shading because the short shoot production of neither the branch order 2 nor the branch order 3 was affected by shading. Suzuki and Maki Suzuki [27] reported that, in Eurya japonica, the annual shoot length decreased with increasing branching order as has been found for many young and adult conifers [11,12,14] like pines and spruce [13].
However, Sabatier and Barthélémy [8] also found that annual final shoot length, duration of extension, and maximum extension rate decreased with increasing branching order.
is study showed that below the annual shoot, the length of all types of axillary short shoots did not significantly vary between branching orders 2 and 3 in the same light condition (Figure 1(c)). On the other hand, the order 3 branches behave differently than the lower order one with regard to the effect of sun exposure on the short shoots SSnr and SS (nr + r) (Figure 1(c)). e increase of the length of the short  Figure 2: Variation of the total number of production and death of short shoot needles: (a) the mean number of fallen leaves of different short shoot needles (b) among L2, D2, L3, and D3 branches exposed to different light intensities as described in Table 1. Different letters indicate only significant differences among branches (P < 0.05, n � 6).  Figure 3: Variation of the mean frequency per branch of short shoots carrying pollen strobili in different light intensities as described in Table 1. Different letters indicate significant differences among branches (P < 0.05, n � 6). Table 3: e mean frequency of sexual short shoot per type of short shoot of the four experimental branches exposed to different light intensities as described in Table 1  shoot SSnr and SS (nr + r) of the order 3 branches in sun exposure shows that when the cedar is more exposed to the sun rays, it reacts by extension of the short shoot of the youngest branches. is response of the order 3 to light availability is supported by the fact that C. libani is found to be a light-demanding species [39] and when growing under better light and water availability showed higher growth rates and wider tree rings [29]. In contrast to our result (Figure 1(c)), Remphrey and Powell [16] and Remphrey and Davidson [15] reported that the length of current shoots of major order 1 branches decreases with decreasing light and the older branches grow slowly, while Suzuki [18] found that Cleyera japonica annual shoot length and light environment were positively correlated within the same order branches and particularly the order 2.
ey also reported that the lower-order 2 branches Cleyera japonica lengthened to a certain extent even if light availability was relatively low as in this study for branch order 2, whereas they observed that higher-order 3 and 4 branches did not grow vigorously even when light availability was relatively high. Our observations with those of Suzuki [18] are not corroborated with the observations of Remphrey and Powell [16] and Remphrey and Davidson [15] and Chen and Sumida [40] who found that for the primary and secondary branches, the mean shoot lengths in the lower crown were not significantly different between sunlit and shaded trees of Sakhalin Spruce. e low shoot length growth of a shaded order 3 branch compared to sunlit one obtained here is due to correlative inhibition and sink/source competition between sunlit and shaded branch for the partially shaded tree resources as explained by Milton's law of resource availability and allocation [35,[41][42][43][44]. Moreover, the observation that sunlit branches, as in the case of order 2 branch, cannot usually grow longer than shaded branches shows that competition between shaded and sunlit branch for the partially shaded tree resources is branch age-dependent and can be tested by branch anatomy analysis as suggested by Umeki and Seino [10]. e impact of light on needle retention depends on the order and the type of short shoot so that the shade increases needle retention with increased order only in SSnr and SS/T short shoots (Figure 2(b)). is action is due to the higher water status of the branch order 3 than the branch order 2 and is in agreement with the data of Suzuki [18] and Grönlund et al. [45] who reported that specific leaf area conductivity of new shoots declined with parent branch age and increased with simulated transpiration rate of the shoot. Furthermore, Jalkanen et al. [46] found a lower decrease of the mean annual summer needle retention in northern than in southern Finland and suggested that a tree's carbon balance may better explain the variation in needle retention.
is explanation is in agreement with the idea of shade preserves branch water level by stimulation of stomata closure and reduction of water use in photosynthesis and growth as observed by Robakowski et al. [47].
In the order 2 branches, we found that only short shoot SS (nr + r) that does not lengthen significantly more under the effect of sun exposure loses significantly more needles compared to shade. Contrarily in the order 3 branches, only short shoot SSnr that elongates significantly more under the effect of sun exposure loses more needles whereas SS/T that does not extend more under sun exposure also loses their needles more significantly compared to shade. ese results show that in the proximal area of the adult cedar crown, the short shoot SS/T of the young branches behaves in the same way as the short shoot SS (nr + r) of the older branches in terms of extension and needle loss. ey also show close relationships between short shoot extension and needle life span of axillary products in the proximal part of C. atlantica tree crown.
Our results show that the light conditions affect the needle life span of the length of the short shoot particularly for order 3 branches. e main response to the shadow here is the length of unbranched short shoot suggesting that the nature of short shoot may vary between vegetative state and extremely specialized state with a role of "mega leaf" carrying the sexual organ (male or female strobili). is observation suggests that the shadow may be suspected to also affect cedar male and female sexual functions carried by the unbranched short shoot as reported by Nuñez et al. [48] for Austrocedrus chilensis (D. Don) who found that male trees were more abundant on high-radiation slopes, while lowradiation slopes had higher abundances of females. On the other hand, the effect of shade on the needle life span of the length of unbranched short shoot was higher in young than in old branches (Figures 1(c) and 2(b)), while the effect of shade on the production of male strobili was higher in old than in young branches ( Figure 3 and Table 3). We suggest that this is an index of the plant actively extending its canopy towards a preferred light environment, temporarily giving up of functions less vital than vegetative growth.
As regards reproduction, the pollen strobili production per branch was about five times more in the sunlit order 2 branches (L2) as compared to shaded order 2 branches (D2).
is result is in agreement with data of Henriksson [38] who has found that the nonshaded branches of the partially shaded trees of Scots pine grew well and produce more male catkins than the shaded branches of the same tree because sunlit branches are believed to be strong sinks for water and carbon on partially shaded tree [44]. For example, if within the sunlit order 2 branches which are nonshaded branches, the net photosynthesis and the stomata conductance are increased after shading the lower branches, in this case shaded order 2 branches (D2), this will, in long term, inevitably lead to increases of the production of sexual short Table 4: ANOVA (P < 0.05) of the effect of order and light availability on the mean frequency of sexual short shoot produced per type of short shoot in experimental branches exposed to different light intensities as described in Table 1 International Journal of Forestry Research shoot at the costs of the shaded branches. is competition for resources favours sunlit L2 branch reproduction while that between L3 and D3 branches favours sunlit L3 branch growth and is in accordance with the sink/source competition. Henriksson [38] believed it to play a role in determining where the growth and reproduction occurred in the canopy.
On the other hand, in this study, both sunlit branches of orders 2 and 3 were found to be relatively similar in their growth and reproduction whereas both shaded branches of orders 2 and 3 are similar in reproduction but differed in their growth in comparison with sunlit branch of the same order (Figures 1-3). e shaded branches D2 are similar to sunlit L2 one in growth while the shaded branches D3 grew less than sunlit L3 one. is indicates that both sunlit branches which are strong sink branches drained more resources to the unbranched short shoot, at the cost of the growth and reproduction of shaded branch. is observation is supported by the fact that growth of short shoot of the old sunlit branches is not affected by shade while their sexual short shoots are more decreased and vice versa for the young branch ( Figure 3 and Table 3).
is means that the old branches have a much larger pool of carbohydrates to allocate throughout their short shoots because they become able to compensate for the growth expenses it has in shade but in sunlight fix enough carbon to repay pollen strobili construction costs. Many studies have showed that differential reproductive effort between sexes that is not compensated for may translate into different vegetative growth rates [49][50][51][52][53][54][55][56][57] also support this conclusion. As evidenced by ring widths, Nuñez et al. [48] have found that female trees of Austrocedrus chilensis were able to compensate for the reproductive expenses they have only in low-radiation exposures whereas in this study Cedar partially shaded tree is not able to compensate for the production of male strobili costs it has in shaded branch.
It has been reported that the productivity, the interval of production, and the release of pollen strobili depend on year, population, and species [58,59] while for Atlas cedar Smith [60] noted "irregularity" with the development of pollen and Bell et al. [32] found no significant relationships between climate, soil nutrient availability, and pollen size. Besides buffering against these unpredictable factors, we argue that the age affects the branch sexuality sensitivity of Cedrus to shade and must be considered at the individual tree level.
is is of immense value particularly for tree breeders because pollen strobili availability is a major constraint of tree reproductive success. Furthermore, other studies must be done on female strobili of different crown parts to better understand the effect of environmental change on the reproduction of Atlas cedar which is fundamental to the conservation of Cedar forests in Morocco.

Data Availability
e data used to support the findings of this study are included within the article.

Conflicts of Interest
e authors declare that they have no conflicts of interest.