Prioritizing conservation seed banking locations for imperiled hemlock species using multi-attribute frontier mapping

Hemlock woolly adelgid (Adelges tsugae Annand) (HWA) is an invasive forest insect sweeping across the native range of eastern (Tsuga canadensis [L.] Carr.) and Carolina (Tsuga caroliniana Engelm.) hemlocks, threatening to severely reduce eastern hemlock extent and to push Carolina hemlock to extirpation. HWA poses a significant threat to these eastern US natives, now infesting hemlocks across 19 states and more than 400 counties. For the long-term preservation of the species, ex situ genetic resource conservation efforts such as seed collection, storage, and adelgid-resistant hemlock breeding may all be necessary. To ensure the efficient and effective application of these efforts, it will be necessary to prioritize locations within the native ranges, because it is logistically impossible to apply these efforts to all populations. To build upon 12 years of seed banking for eastern and Carolina hemlock, we applied a novel approach for incorporating multiple dissimilar data sets into a geographic prioritization of areas for the most effective and efficient conservation of genetic diversity. The approach involves integration of geographic information systems with the multi-attribute frontier mapping technique to identify locations across the ranges of these two imperiled species most in need of conservation actions. Specifically, our assessment incorporated four genetic diversity parameters, a climate component, a measure of population disjunctiveness, a measure of local hemlock abundance, and seed collection density to prioritize areas of eastern and Carolina hemlock occurrence for ex situ gene conservation. For each species, the result was a mapped index of locations prioritized by the combined significance of these factors. For eastern hemlock, this assessment assigned the highest priority to disjunct populations and to some areas within the main-body range. With Carolina hemlock, disjunct northern populations and central main-body locations received the highest prioritization. Our prioritization approach could be applied similarly to other species facing pressure from invasive pests or other environmental threats.     

Response of arthropod biodiversity to foundation species declines: The case of the eastern hemlock

Widespread declines of foundation species, such as many corals, kelps, and overstory trees, are of grave concern because, by definition, these species create and maintain habitat that supports other species. Nevertheless, past responses to their declines, many of which were caused by invasive species, have been late and ineffective, underscoring the need to predict changes in biodiversity and ecosystem function associated with species invasions and foundation species losses. One predictive, but under-used, approach is to compare the species and functions associated with the afflicted foundation species to its projected replacement communities. The taxa associated with the foundation species and subsequent successional stages would be expected to decline and increase, respectively. We used this approach to generate hypotheses for how arthropod diversity might change in response to extensive losses of eastern hemlock trees caused by the invasive, hemlock woolly adelgid (insect: Hemiptera, Adelgidae). Our all-strata survey of the arthropods in an eastern hemlock forest and its expected replacement climax community in the mid-Atlantic region of the United States, mixed hardwood forest, suggests that eastern hemlock losses might initiate increases in arthropod abundance, alpha diversity, and 23 arthropod taxa, might produce no change in evenness or composition of arthropod functional groups, but might trigger decreases in beta diversity and seven hemlock indicator taxa. These predictions are consistent with observed trends in arthropod responses to hemlock losses in other studies, and thus might be useful for targeting early monitoring, management, and conservation efforts. This research is exploratory, however, and tests of these predictions across larger spatial scales will be necessary to determine the generality of the findings.     

An analysis of the pattern of genetic variation in Vitellaria paradoxa using RAPD markers

Vitellaria paradoxa C.F.Gaertn. is one of the most economically and socially important tree species in the Sudano-Sahelian region. Little is known of the pattern of variation within its natural range. Eight populations covering most of the natural range from Senegal to Uganda were sampled and leaves of 118 individual trees were collected. An analysis of molecular diversity was carried out using random amplified polymorphic DNA (RAPD) markers. Fifteen random primers generated 67 polymorphic and 15 monomorphic RAPD loci ranging from size 1670 bp to 280 bp. Shannon's diversity index varied from Central Africa/Ndele (0.374) to Uganda/Amoya (0.350) but the differences between populations were smaller than the population standard errors. Correspondence analysis of unrooted neighbour-joining trees suggested that genetic distances between populations were correlated with geographic distances. This trend was confirmed by a Mantel test giving a coefficient of correlation between genetic and geographic distances of R = 0.88 (P = 0.0001). Result of AMOVA (analyses of molecular variance) showed that 14.8% (P = 0.002) of the RAPD variation was distributed among populations. Nested analysis of variance indicated that variance between the western and eastern groups of population represented 8.7% (P = 0.001) of the total variation and the variation amongst populations within group was 9.5% (P = 0.001). Eighty two percent of the variation was explained by variation amongst individuals within populations. The origin of genetic structure and level of diversity may be explained by the glacial refugia, the biological traits of Vitellaria paradoxa and by the impact of semi-domestication. Based on these results, sampling options of the natural populations are suggested for in or ex situ conservation. For the development of Vitellaria paradoxa breeding population, the sampling should consist of many individual trees selected within a few populations to capture a large proportion of variation.This revised version was published online in November 2005 with corrections to the Cover Date.     

Growth and Infestation by Hemlock Woolly Adelgid of Two Exotic Hemlock Species in a New England Forest

ABSTRACT

The hemlock woolly adelgid (HWA, Adelges tsugae Annand) an invasive exotic insect, may extirpate eastern hemlock (Tsuga canadensis (L.) Carrière) trees from native forests, but other hemlock species could be planted to occupy their ecological niche. This study tests two of the most likely replacement species candidates: western hemlock (T. heterophylla (Raf.) Sargent) and Chinese hemlock (T. chinensis (Franchet) Pritzel). Low survival rates, slow growth, and infestation by HWA of western hemlock in eastern hemlock forests shows that the western hemlock is not a likely candidate for planting in the northern portion of eastern hemlock's range. In contrast, Chinese hemlock grew at rates similar to eastern hemlock and did not show any signs of HWA infestation. In this study, damage from deer was a much bigger problem than growth reductions from HWA.     

Genetic analysis and conservation of endangered medicinal tree species <Emphasis Type="Italic">Taxus wallichiana</Emphasis> in the Himalayan region

Taxus wallichiana is one of the most important medicinal tree species of the Himalayan region. Leaf and bark of the species yield an important drug called taxol, which is used for treatment of many types of cancer. There is a serious threat to the existence of the species due to over exploitation in its native habitat. Adequate information on the nature and the extent of genetic diversity in this important species is required for developing suitable strategy for its conservation. Random amplified polymorphic DNA (RAPD) and amplified fragment length polymorphism (AFLP) were used to assess genetic variation in nine natural populations of T. wallichiana from western part of the Himalayan ranges. Both the markers revealed low genetic diversity in these populations. Average heterozygosity for AFLP and RAPD were 0.3715 and 0.3072, respectively. Φ_ST values derived from molecular variance were 0.0855 and 0.1005 for AFLP and RAPD, respectively, whereas the corresponding G_ST values were 0.1796 and 0.2140. Most part of the genetic variation was present within the populations. However, between population variation was low but statistically significant, which suggested that the sampled populations might not constitute a single panmictic population. Cluster analysis and Mantel’s correlation revealed that genetic differentiation broadly followed geographic distribution of the populations. T. wallichiana thus urgently needs to be conserved using both in situ and ex situ conservation approaches.     

Species differences in total and vertical distribution of branch- and tree-level leaf area for the five primary conifer species in Maine, USA

Vertical distribution of leaf area largely governs both tree structure and function. Models of this important tree attribute have been constructed for several commercially important conifers. However, a limited number of studies have compared alternative modeling techniques and inherent species differences. This study used several existing datasets for the five primary conifer species in Maine, namely balsam fir [Abies balsamea (L.) Mill.], northern white-cedar [Thuja occidentalis (L.)], eastern hemlock [Tsuga canadensis (L.) Carr.], eastern white pine [Pinus strobus (L.)], and red spruce [Picea rubens (Sarg.)] to examine species variation in total and vertical distribution of projected leaf area at the individual branch- and tree-levels. In addition, multiple methods for modeling the vertical distribution of leaf area were examined across the species. For a given branch diameter and location within the crown, eastern hemlock branches held the greatest amount of leaf area, followed by balsam fir, northern white-cedar, white pine, and red spruce. At the tree-level, eastern white pine held the greatest amount of leaf area followed by eastern hemlock, balsam fir, red spruce, and northern white-cedar for a given tree size. Across species, the two-parameter, right-truncated Weibull distribution performed the best for predicting vertical distribution of leaf area when compared to the four-parameter beta and Johnson's S_B distributions (reduction of root mean square error of 1.7–21.1%). Northern white-cedar had a relative distribution of leaf area distinctly different than other species in this study with a mode shifted towards the upper crown. In contrast to red spruce and white pine, the mode of the relative distribution of leaf area for balsam fir and eastern hemlock occurred lower in the crown. Results of this study suggest that differences in total and vertical distribution of leaf area exist between species, but significant amounts of their variation are largely accounted for by bole and crown size.     

Genetic diversity and differentiation of yellowwood [<Emphasis Type="Italic">Cladrastis kentukea</Emphasis> (Dum.Cours.) Rudd] growing in the wild and in planted populations outside the natural range

Yellowwood (Cladrastis kentukea) grows in small, widely scattered populations in the wild, but is also a popular ornamental tree that thrives when planted in urban areas outside its natural range. Since the small native populations of yellowwood in several states are considered at risk of extirpation, the cultivated population could serve as an ex situ repository of yellowwood genetic diversity that could be used to restore lost local populations of the species. The potential value of cultivated yellowwood for conservation depends on the genetic diversity among cultivated trees compared to natural populations. Using nuclear microsatellite markers, we genotyped 180 yellowwoods from natural populations in Indiana, Missouri, Arkansas, and Kentucky, along with 61 trees from urban parks and landscapes in Indiana, Ohio, and Missouri. We found that, even when statistics were adjusted based on population size, the urban “population” had higher genetic diversity than any of the wild populations sampled, indicating that commercially-grown yellowwood is most likely a mixture of genotypes from isolated wild populations. We observed strong genetic differentiation among wild populations, and evidence for inbreeding in at least one of the wild populations.     

Assessment of morphological and genetic diversity in <Emphasis Type="Italic">Gmelina arborea</Emphasis> Roxb.

Gmelina arborea is an important timber-yielding tree that grows naturally in the tropical and subtropical regions of Southeast Asia and has also been introduced as a plantation species outside these regions. Genetic diversity in this tree species was observed in stone/seed-related traits and in vitro responses of cultured nodal segments from plants of eight different populations representing natural forests, fragmented forests and plantations. Variance analysis showed significant differences between populations for these traits. However, it was not possible to separate the different populations using these traits by multivariate analysis, even after environmental variation was reduced over six subcultures. Genetic diversity was therefore analysed using molecular markers. Inter-simple sequence repeat (ISSR) primers yielded 95% polymorphic loci among the eight populations and UPGMA analysis enabled separation of these populations on the basis of their genetic distances. Diversity was also analyzed using population genetics parameters like Nei’s genetic diversity and gene differentiation. Nei’s genetic diversity was 0.29 between populations and 0.11 within populations. AMOVA analysis indicated 41 and 59% within- and between-population genetic diversity, respectively. Mantel test suggested that genetic differentiation between six Indian populations was positively correlated to geographic distance (r = 0.626, P = 0.029). Assessment of the genetic variation in G. arborea populations is an important step in selection of conservation strategies for this species since diversity forms the basis for species adaptation.     

Identification of salt tolerant baldcypress (<Emphasis Type="Italic">Taxodium distichum</Emphasis> (L.) Rich) for planting in coastal areas

Coastal baldcypress (Taxodium distichum) forests are being killed in the southern United States as a result of saltwater intrusion. Tank studies have indicated substantial intraspecific variation in salt tolerance within baldcypress populations, but only one field study has been conducted and it used a limited seed source. The major objective of this project was to expand the collection range of baldcypress across the southeastern United States and determine if there are baldcypress populations that can survive and grow in saltwater damaged areas. Seeds were collected in 1996 from eight estuarine areas (James River = VA; Cape Fear River = NC; Winyah Bay = SC; Ogeechee River = GA; Ochlockonee River = FL; Mobile Bay = AL; Biloxi River=MS; Chalmette = LA). Seeds were stratified and planted, and seedlings were grown for 2 years before planting in two abandoned ricefields on Hobcaw Barony near Georgetown, South Carolina in 1999. Salinity levels reached 18.5ppt during 2001 at the peak of the worst drought on record. By 2002, the only seedlings surviving were those from LA, AL, and FL. After 5 years in the field, LA seedlings were the best performers with a mortality rate of only 27%. More detailed analyses of the biologic and genetic characteristics of trees growing in the LA site need to be completed to determine if they represent a source of baldcypress seed possessing greater tolerance to saline conditions. If so, seeds need to be collected and nurseries established to grow seedlings to help restore degraded wetland swamp areas along the Gulf and Atlantic coasts.     

Hydraulic responses to environmental perturbations in Tsuga canadensis and Betula lenta

Eastern hemlock (Tsuga canadensis (L). Carr.) is a late-successional species found across the northeastern United States of America that is currently threatened by the exotic pest, hemlock woolly adelgid (Adelges tsugae Annand). Because whole-tree physiological characteristics may scale to influence ecosystem processes, we considered whole-tree hydraulic controls in eastern hemlock and the replacement species black birch (Betula lenta L.). Through a series of misting perturbations, whole-tree resistances (R), capacitances (C) and time constants (tau) were determined from time series sap flux data in eastern hemlock and black birch. Black birch trees responded more rapidly to environmental perturbations than eastern hemlock. Utilizing the step function after applied treatments, whole-tree tau ranged between 9.4 and 24.8 min in eastern hemlock trees compared with 5.9 to 10.5 min in black birch. Species was not a significant predictor of R or C when controlling for tree size. In both species, R decreased with sapwood area and C increased. Our tau results indicate that the loss and replacement of eastern hemlock by black birch will decrease the lag between transpiration and absorption of water from the soil and potentially alter the diurnal pattern of carbon and water uptake.     

Conifer species influence Heterobasidion irregulare sporulation on stem discs exposed to natural inoculum and root wood decay in soil‐block tests

The fungal pathogen Heterobasidion irregulare has caused substantial damage to pine species in much of the south‐eastern United States, but the potential for damage from Heterobasidion root disease to forest tree species of the Great Lakes Region has not been thoroughly investigated. In each of three trials in two infested stands, fresh stem discs of six native conifer species were exposed to natural inoculum and examined for the asexual Spiniger stage of H. irregulare after incubation in the laboratory. In the first trial, the fungus was infrequently observed on discs, and differences among species were not found. But in the other two trials, detection of the pathogen was much more frequent on discs of tamarack (Larix laricina), red pine (Pinus resinosa) and white spruce (Picea glauca) than those of northern white cedar (Thuja occidentalis), eastern hemlock (Tsuga canadensis) and balsam fir (Abies balsamea). In a second study, the ability of an isolate of H. irregulare to decay root sapwood of these same species was compared using a soil‐block decay test. After 16 weeks of incubation, weight loss due to decay was greatest for red pine (23.2%) and larch (19.1%) and least for balsam fir (1.3%). Differences in the ability of H. irregulare to infect and saprophytically utilize substrate of these important tree species should prompt additional investigation of the relative risk posed to forests of the Great Lakes Region.     

Changes in light levels and stream temperatures with loss of eastern hemlock (Tsuga canadensis) at a southern Appalachian stream: Implications for brook trout

The exotic invasive insect, hemlock woolly adelgid (Adelges tsugae Annand), is causing mortality in eastern hemlocks (Tsuga canadensis [L.] Carr.) throughout the eastern U.S. Because hemlocks produce dense shade, and are being replaced by hardwood species that produce less shade, their loss may increase understory light levels. In the southern Appalachians, increases in light could increase stream temperatures, threatening species such as brook trout (Salvelinus fontinalis). We studied changes in light and stream temperature with eastern hemlock decline at a headwater southern Appalachian brook trout stream. Our results indicate that stream light levels have increased significantly with adelgid infestation. Leaf-on light levels are currently significantly higher (P < 0.02) in plots containing high basal areas of hemlock (mean global site factor (GSF)(SE) = 0.267(0.01)) compared with plots containing no hemlock (mean GSF(SE) = 0.261(0.01)), suggesting that increases in light have occurred with hemlock decline. The Normalized Difference Vegetation Index (NDVI), a remotely sensed metric of vegetation density, decreased with hemlock decline from 2001 to 2008. In 2001, NDVI showed no relationship (R² = 0.003; F = 0.14; P = 0.71) with hemlock basal area, but by 2008, there was a significant negative relationship (R² = 0.352; F = 19.55; P < 0.001) between NDVI and hemlock basal area. A gap experiment showed that light levels may increase by up to 64.7% more (mean increase in GSF = 27.5%) as hemlocks fall, creating gaps in the canopy. However, stream temperatures did not increase with hemlock decline during the study period, and we found that ground water inputs have a stronger influence on water temperature than light levels at this site. Linear regression showed a significant negative relationship between water temperature and proximity to ground water sources (R² = 0.451; F = 13.14; P = 0.002), but no relationship between water temperature and light levels (R² < 0.02; P > 0.05). In addition, by comparing light levels between plots containing hemlock and those containing only hardwoods, we found that if hemlocks are replaced by hardwoods, light levels under an all-hardwood canopy (mean GSF(SE) = 0.240(0.005)) are unlikely to be higher than they are under the current forest (mean GSF(SE) = 0.254(0.007)). These results suggest that loss of hemlock along southern Appalachian headwater streams could have short-term impacts on light levels, but that long-term changes in light levels, increases in water temperature, and adverse effects on brook trout may be unlikely.     

云南松主分布区天然群体的遗传多样性及保护单元的构建

以20个云南松天然群体为对象,SSR分析表明云南松遗传变异主要存在于群体内,各群体间的遗传多样性水平差异不显著。基于SSR分子标记遗传多样性分析,根据Nei's遗传相似度,采用逐步聚类优先取样法,分别对初始群体、遗传多样性保护单元和剩余群体进行比较、t检验,以此评价遗传多样性保护单元的代表性。结果表明,利用该方法抽样的群体能很好地代表初始群体的遗传多样性,按30%的群体抽样,等位基因数、有效等位基因数、Shannon's信息指数和期望杂合度的保留率分别达到初始群体的98.03%、105.36%、103.99%和105.56%。保护单元的遗传多样性均高于剩余群体,抽样群体组成的遗传多样性保护单元的效果较好。     

Response of soricid populations to repeated fire and fuel reduction treatments in the southern Appalachian Mountains

Fuel hazards have increased in forests across the United States because of fire exclusion during the 20th century. Treatments used to reduce fuel buildup may affect wildlife, such as shrews, living on the forest floor, especially when treatments are applied repeatedly. From mid-May to mid-August 2006 and 2007, we used drift fences with pitfall traps to capture shrews in western North Carolina in 3 fuel reduction treatment areas [(1) twice-burned (2003 and 2006), (2) mechanical understory cut (2002), and (3) mechanical understory cut (2002) followed by 2 burns (2003 and 2006)] and a control. We captured 77% fewer southeastern shrews (Sorex longirostris) in mechanical + twice-burned treatment areas than in mechanical treatment areas in 2006, but southeastern shrew captures did not differ among treatment areas in 2007. Total shrew captures did not differ among treatment areas in either year. Decreases in leaf litter, duff depth, and canopy cover in mechanical + twice-burned treatment areas may have decreased ground-level moisture, thereby causing short-term declines in southeastern shrew captures. Prescribed fire or mechanical fuel reduction treatments in the southern Appalachian Mountains did not greatly affect shrew populations, though the combination of both treatments may negatively affect some shrew species, at least temporarily.     

Variation in seed and seedling traits among natural populations of <Emphasis Type="Italic">Trigonobalanus doichangensis</Emphasis> (A. Camus) Forman (Fagaceae), a rare and endangered plant in southwest China

The extent of genetic variation in seed and seedling traits of Trigonobalanus doichangensis, a rare and endangered tree in southwest China, was evaluated at the population level. Seeds were collected from four natural populations in the south and southwest Yunnan province, China. Significant differences (P < 0.05) between populations were observed in seed length, seed weight and seedling morphological characters. The analyses revealed that none of the seed and seedling traits had significant correlations with the geo-climatic variables of the populations. Among seed and seedling traits, significant correlations (P < 0.05) were discovered among seed length, seed weight and height of 30-day-old seedlings. Significant correlation (r = 1.00, P < 0.01) occurred also between the height and root collar diameter of 4-year-old seedlings. With the exception of seed width and seed germination percentage, 51–96% of total variation in other seed and seedling related traits was attributed to population effect, as determined by the broad sense heritability. In conclusion, these observations on variation in seed and seedling traits should be a reference point, when considering seed collection of this species for ex situ conservation and species restoration.     

Phylogeography and conservation perspectives of an endangered macaronesian endemic: Picconia azorica (Tutin) Knobl. (Oleaceae)

The Azores provide an excellent field test for research activities aimed at developing conservation strategies for endangered tree species. In this work, the urgency to promote Picconia azorica conservation programs addressed (a) insights into the biotaxonomy of the species (including an evaluation of its origin and relationships with the only congeneric species of P. azorica: P. excelsa from the Canary and Madeira islands) and (b) the evaluation of P. azorica genetic diversity. Plastid DNA sequence analysis and molecular markers (RFLP and SSR) were used for this purpose. Phylogenetic data suggest the monophyly of Picconia and support a late Miocene divergence of the two species. Three polymorphic cpSSR loci allowed the identification of five different haplotypes in P. azorica. Uniqueness and relictuality of lineages are presented and discussed. Picconia azorica intra-specific diversity patterns revealed low genetic diversity and a weak genetic structure, which could result from long-lasting ecological stability and efficient inter-island seed movement that have been severely affected in recent times. The species survival is at risk, and we suggest management practices focusing on ex situ and in situ conservation units based on eco-genetic data. Additional measures contributing to mild erosion of the genepool and to remove barriers to seed dispersal are indicated.     

森林遗传资源保存的现状和策略

综述了世界森林遗传资源保存的历史和现状,作者认为中国森林遗传资源的保存策略应为系统管理、有效保存、合理利用、丰富发展。     

Genetic diversity analysis in a seed orchard of <Emphasis Type="Italic">Eucalyptus tereticornis</Emphasis>

India is the largest grower of Eucalyptus with an area of 3.943 million hectares under plantations and E. tereticornis is the predominant species in the plains of southern India with an average productivity of 12–25 m³ ha⁻¹ year⁻¹. With the aim to establish seedling seed orchards of the species, seed lots of fifteen provenances were imported from Australia and a trial was laid. In the present study the genetic diversity existing in the seed orchard was estimated using ISSR–PCR. Seven ISSR primers amplified 663 amplicons in the size ranging from 255 to 2,711 bp. The total number of polymorphic bands varied from 59 to 123 with 100% polymorphic banding profiles. The average gene diversity (Hj) of all the provenances ranged from 0.0589 to 0.1109 and the total gene diversity estimated was low (H _T = 0.130) when compared to the earlier reports from other eucalypts species. Analysis of Molecular Variance partitioned the ISSR variation into inter and intra population components. The inter population component accounted for 55.2% of the variation and the intra population component accounted for 46.3% (P < 0.001). A neighbour-joining analysis was done using the dissimilarity matrix to determine the aggregation of the individuals into clusters. Existence of population structure among the populations was revealed in STRUCTURE analysis but geographical region based clustering was not observed. The assessment of intra and inter genetic variation documented in the present study suggests that, along with the phenotypic traits, knowledge about genetic diversity measured at the DNA level in individuals of seed orchards provide an objective guide for selective culling of trees for maintaining optimal diversity for enhanced genetic gains.     

Genetic diversity in Tunisian Crataegus azarolus L. var. aronia L. populations assessed using RAPD markers

? The genetic diversity of nine wild Tunisian Crataegus azarolus var. aronia L. populations from different bioclimates was assessed using RAPD markers.

? Eight selected primers generated a total of 105 bands, 81 of which were polymorphic. Shannon’s index (H′) ranged from 0.222 to 0.278 according to a population with an average of 0.245. The genetic variation within the species (H _SP = 0.423) was relatively low. A high differentiation (G _ST = 0.421) among populations coupled with a low level of gene flow (N _m = 0.472) were observed. The analysis of molecular variance (AMOVA) revealed also significant differentiation among populations (Φ_ST = 0.371), even at a low scale space. The majority of variation occurred within populations (63.31%). The Mantel test performed on genetic (Φ_ST) and geographic distance matrices among population pairs did not reveal an isolation by distance.

? Interpretation of Neighbour-joining tree based on Nei’s and Li’s genetic distance among individuals showed distinct population groupings. The UPGMA dendrogram based on Φ_ST values revealed two population sub-clusters, each including populations from different bioclimates and/or geographic regions.

? The low level of genetic diversity and the high genetic structure of populations resulted from genetic drift caused both by habitat fragmentation and the low size of populations.

? The high differentiation among populations and the similar low level of diversity within populations suggest that in situ conservation should interest all populations. The ex situ conservation should be based on the collection of seeds rather within than among populations because of the maximum of variation was revealed within populations.