Click for next page ( 52


The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement



Below are the first 10 and last 10 pages of uncorrected machine-read text (when available) of this chapter, followed by the top 30 algorithmically extracted key phrases from the chapter as a whole.
Intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text on the opening pages of each chapter. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

Do not use for reproduction, copying, pasting, or reading; exclusively for search engines.

OCR for page 51
Structure of Genetic Variation One of the most fundamental requirements for improved man- agement of forest tree diversity is understanding the biological dynamics of genetic variation within and between tree species. Information about the genetic architecture of tree populations will make it possible to develop sounder strategies for their conservation and use. MATING SYSTEMS AND GENE FLOW Considerable variation exists among tree species with respect to the extent of genetic diversity and the way such diversity is spatially or temporally organized within and among populations. The extent and the pattern of genetic diversity in forest trees are strongly regulated by their mating patterns and gene flow. Despite decades of research on the reproductive biology of forest trees, knowledge about the subject is still limited. Mating Mechanisms Mating systems among trees are quite varied. Most tropical forest trees have hermaphroditic flowers; that is, the flowers contain male and female reproductive organs (Ashton, 1969; Bawa, 1974, 1979~. In contrast, a vast majority of temperate tree species are monoecious; that is, they are characterized by the presence of unisexual flowers, but male and female flowers occur on the same plant. Dioecy (the presence of separate male and female plants) occurs in temperate and tropical species, but 51

OCR for page 51
52 / Forest Trees it is more common in the latter (Bawa and Opler, 1975~. The evolutionary dynamics of forest trees is further complicated by the existence of mixed mating and reproductive systems. Selfing (self-pollination or self-fertil- ization) is possible in hermaphroditic and monoecious species, but it is usually prevented by a wide variety of genetic mechanisms or differences in the maturation of male and female floral parts. Many tropical species are genetically self-incompatible; that is, little or no seed is set following self-pollination (Bawa, 1974; Bawa et al., 1985~. Surprisingly little is known about self-incompatibility in temperate species with hermaphroditic flowers; in the monoecious species, two mechanisms prevail that prevent or reduce selling. First, the maturation of male and female flowers at different times, as happens, for example, in most temperate coniferous trees, reduces the possibility of selling. Second, self-sterility, which also occurs in most conifers, causes most selfed seeds to be aborted before they mature. This mortality is assumed to be due to lethal recessive genes that are brought together when individuals that are closely related genetically are crossed (inbreeding). In sum, designing tree reserves and managing natural or artificial stands of trees requires an understanding of the reproduction of the trees involved. Without it, tree populations could, for example, fail to reproduce, or they could experience excessive or unintended inbreeding and might thus become endangered. Inbreeding Various mechanisms can eliminate the possibility of selling, but they do not exclude inbreeding. The level of inbreeding is determined not only by the nature of the reproductive system, but also by family structure, which itself is influenced by pollen and seed dispersal characteristics. Dispersal of pollen and seeds over a limited area can increase the genetic relatedness of nearby individuals in a mating population and, thereby, the potential for inbreeding. The longevity of some trees over others in the population can also result in a relatively few individuals genetically dominating the gene pool. This, combined with such factors as asynchronous flowering of male and female flowers, overlapping generations within the populations, and unequal sex ratios, can potentially increase the level of inbreeding. Apomixis, or uniparental reproduction, which appears to be quite common in some tropical trees (Ashton, 1988), may also contribute to inbreeding. The net result of these factors is that the effective population size in terms of reproductive capacity is often much smaller than the total number of adult trees.

OCR for page 51
Structure of Genetic Variation 153 Outcrossing Outcrossing refers to the mating of genetically nonidentical individ- uals. The rate of outcrossing is often estimated by the use of genetic markers in the form of allozymes (genetic variants of enzymes). In temperate and tropical trees, outcrossing rates have been found to be very high. Although the outcrossing rate can range from 0 (no outcross- ing) to 1 (100 percent outcrossing), it varies from 0.60 to 1 in the tree species examined by Knowles et al. (1987), and wide variations in rates occur among individuals even of the same population. The distances between trees and the timing of reproductive flowering (i.e., spatial and temporal variation) can also affect outcrossing rates. Rates in tamarack, for example, have been found to vary with stand density, the outcrossing rate being lower in less dense than in more dense stands. A practical implication of this result is that spatial isolation of trees due to forest decline or degradation may increase the level of inbreeding and its concomitant effects. Gene Flow Gene flow implies an exchange of gametes or genes among dispersed trees, and it is inversely related to population differentiation. It occurs through the movement of pollen and seed, which can be dispersed by a wide variety of nonliving and living mechanisms. Knowing how populations are expanded, maintained, or restricted by gene flow is essential for managing distinct tree populations in their original envi- ronment. Gene flow determines the geographic scale over which populations may be differentiated from each other. Population differentiation can also occur in response to selection resulting from local variations in environmental factors. Discriminating between selection effects and migration-induced patterns, however, is difficult. Moreover, selection or sexually divergent migration, which can be frequent among tree species, can have significant impact on gene flow (Gregorius and Namkoong, 1983; Namkoong and Gregorius, 1985~. Where lack of gene flow occurs, the potential for future development of distinct populations exists. All temperate conifers and many hardwood species, such as oaks and poplars, are wind pollinated. Although most pollen is dispersed by windfalls near the paternal tree, much is still carried farther away, several hundred meters or more (Levin and Kerster, 1974~. This, in

OCR for page 51
54 / Forest Trees combination with the high densities of these temperate, wind-pollinated species, results in a potentially high frequency of pollen exchange. Some temperate trees are pollinated by animals (including insects), but virtually nothing is known about the extent to which animal pollinators move among trees. In fact, in most instances even the pollen vectors (carriers) have not been identified. Moreover, in a few instances in which both bird and insect pollinators are effective, as in Camellia japonica (Japanese camellia), the relative visitation frequencies of the pollinators are not known, but it is known that the distances over which they distribute pollen can vary considerably. The diversity of insect and other animal pollinators in the tropics is immense, ranging from bats with a wingspan of 1.5 m to tiny wasps that are 1 or 2 mm in size. Many species, however, are pollinated by medium-sized to large bees (Bawa et al., 1985~. Other pollen vectors are moths, beetles, and the generalist insects, which are the wide variety of small bees, beetles, butterflies, wasps, and flies that visit flowers. THE PACKAGING OF GENETIC INFORMATION A gene, in the classical sense, is the basic unit of heredity and has one or more specific effects on an organism. Genes are segments of DNA (deoxyribonucleic acid) in the nucleus of the cell, linearly arranged to form threadlike structures called chromosomes. The term "locus" refers to the position of a gene on the chromosome, and it is sometimes used interchangeably with the term "gene" when referring to regions of DNA that are influencing a trait. Alternate forms of a gene found at the same locus are called alleles. Some genes have many alleles, which allow for multiple gene products and therefore multiple phenotypes. For example, multiple alleles have been identified for many of the genes that code for human blood proteins. The allelic frequency refers to the proportion of loci in the population occupied by each allele. A gene for which there is more than one common allele is called polymorphic. Most organisms are diploid, meaning they carry two copies of each gene. If both copies are the same allele, the individual is said to be homozygous; if the two copies are different alleles of the gene, the individual is heterozygous for that locus. DNA is the chemical that carries genetic information in all living organisms. The DNA molecule itself is an elongated double helix, often compared with a long, twisted ladder. Corresponding to the rungs of the ladder are two bases, forming a "base pair." The sequence of base pairs at a given locus confers the specificity required for transmission of

OCR for page 51
Structure of Genetic Variation / 55 Relatively few tree species are pollinated by bats and birds, although some important exceptions are the durian fruit (Durio) of Asia, the timber genus Caryocar of the Amazon, many tropical legumes, andspecies of the Bombacaceae. However, these animals do pollinate a large number of other species in the forest. The extent to which pollen vectors move pollen between plants largely depends on their foraging behavior and the spatiotemporal distribution of flowers. Insects and other animals may bring about extensive exchange of pollen among individuals scattered over a wide area (Regal, 1977~. This is particularly true where specificity (uniqueness) between pollen vector and the plant is high. An example is the genus Ficus (fig), in which about 800 species are pollinated by different species of wasps (Wiebes, 1979~. The high specificity enables the fig trees to reproduce and persist even in very low densities (lanzen, 1979~. However, this extreme specificity between pollen vector and plant is not common among tropical forest trees. information via the genes. Due to variability of the base pairs, the number of different alleles that could theoretically be formed from even a very short piece of DNA is extremely large. For instance, a segment of DNA with only 10 base pairs could have more than 1 million different codes. The "average" gene is thought to have up to a thousand base pairs; thus, the DNA structure provides for a phenomenal amount of variation in the genetic code. The difference between two alleles is often as simple as a substitution of a single base pair, but this may correspond to a significant difference in the phenotype resulting from those alleles. Variation can readily be observed at several levels: between species, between major types within a species, between populations within a major type, and between individuals. An individual's genetic composition, or genotype, in conjunction with the environment in which that individual is found naturally, determines the phenotype or observable character . . 1S lCS. Certain traits are controlled by a single gene and are referred to as qualitative or Mendelian traits. Many characteristics, however, are influ- enced by a larger number of genes; these are called quantitative or polygenic traits. The cumulative action of these genes influences the expression of the trait, but the effect of any single gene is small and cannot generally be isolated in the phenotype. Important production traits, such as yield, growth rate, and straightness, are examples of polygenic traits. Occasionally, a "major" gene, one that has a stronger influence on the trait, can be detected, but there is nevertheless modifi- cation of the trait by other genes with smaller effects. ~_ ~

OCR for page 51
~ / ~ ~ ~ ~ ^~ ~ ^ ~ The Quit of F~f~~ If has a ~ed-~ntainin~ stone en- ~sed in ~ white sweet pulp that is att=~cti~ to forge bins and hats. Credit: Douglas ~13. Lit~eis known about the distances aver ibich ani~n~alvectors Wave pollen. Terdiodal binds, such as h~dngLirds; age more sedentary than nonte~dfor~1 ones, and con~sequentl~v, their!po~l~l~e~n~d~ispersal~} extra me~lylimi~d (Linha~,1973~. Bats/ on! The other Bia~nd~,Jre known to ~ awe over a range of seve~~l~kilo~e~ters <~Heiih;~s et~a~l~., 1973) .mI;a~y/ ~eol~-~s~lzec ~! large bees can He pope among plants that am sevefsI~k[~o~ete~ apart (Fr~nkie et<1.,1~976;~J OCR for page 51
Structure of Genetic Variation / 57 Neutral Alleles and the Quantification of Gene Flow Alleles are the alternate forms that a gene may take, which may cause selectively important differences to exist among individuals. Neutral alleles provide no selective advantage or disadvantage and, therefore, are useful for modeling gene flow in populations. The flow of neutral alleles between populations can be quantified by estimating the value of Nm (N = effective size of the local population; m = average rate of gene migration by means of pollen and seeds). The principal reason for estimating Nm is that its value has been shown to indicate the relative importance of gene flow and genetic drift in explaining the observed patterns of genetic differentiation in a population (Slatkin, 1987~. If Nm is less than 1, then changes in allele frequencies resulting from genetic drift of neutral alleles can occur. Such changes are not likely if Nm is greater than 1. As measured by electrophoretically distinguishable alleles, which are often considered to be neutral variations, the values for Nm in a range of forest tree species indicate fairly high levels (> 1.0) of gene flow (Table 3-1~. This is to be expected because most trees have high outcrossing rates, and there is a positive association between outcrossing rate and the level of gene flow (Govindaraju, 1988a). Wind-pollinated species have higher outcrossing rates than animal-pollinated species (Govindaraju, 1988b). The effects of seed dispersal on levels of gene flow have not yet been quantified (Hamrick and Loveless, 1987~. Nonrandom assemblages of genotypes impose a structure on the TABLE 3-1 Relation of Levels of Gene Flow (Nm) to Pollination Mechanism in Selected Forest Tree Species Number of Populations Pollination Species Sampled Nma Mechanism' Abies balsamea 4 4.546 W A. Iasiocarpa 3 7.300 W Berthollefia excelsa 2 1.626 A Eucalyptus caesia 7 0.001 A ssp. caesia E.caesia 6 0.174 A ssp. magma E. cloesia~a 17 1.081 A E. delegates is 8 0.606 A E. oblique 4 0.438 A (continued)

OCR for page 51
If TABLE 3-1 s(O~f~ Number of Populates ~P6llina~S~= Species ampler \ ~{chanis~' S. ~# 3 5. ~A 10 ~810 If 3 1.^ , . . . ,. . . P ~3 1 ~W ~ ~71 3 6 ~W a, . ~ . . ~ ~ ~ ~. ,~ ~41 ~ W ~ ~ ~ ~ - ~ ~ ~ / ~ a, W P In 94.62)W Sap. ~f~- 3. = ~57.~0W . . .. .,^ . ,^ ~^.,^ at, ,, ;~E; ;/ ~; ~t 1 . ~ ~ a. ~5 4.# ~ P. fife ~Tag W UP. ~7 0.051 ~ POW asp. If P. ~5 1.0 ~W . , , en- ^ Off : P. ~ ~6 2.9 ~W P. ~5 l.p ~W 11 8.# , P - '7 ~14 7 483 W .. go, ..... P ~2 ~Z.780 W P. ~2 2.~1 TWO ~W ............ P. / ~1~0 3.020 W 6 S~1 W Q~s~~ 3 2.e5 W Q. j7 ~3 2.4 ~W ^ .. ~. - ~... # ~ I ~ ala v~ . ~ ~ , .~ , ~. .~ a. If 2 3.614 W Q. r ~2 0.< ~ an. ~a 0 5 ~ Is ~2.1 ~W ~ ~ ~ ~ ~ = of pollen and If. Valises of \~ less than 1 indicate the potential far genetic dog. W = mind pll~i~nat~; ^ ~ p~llina~d by an inflect or other animal. u~cE Adapted from D. R. Govindar~iu. 193. Rel~t~i<~nship Street dispersal ability and levels of Cane COD in p)an~. Oikos 52:31~5. Repdnted with permission.

OCR for page 51
Structure of Genetic Variation / 59 population, which can be measured as the departure from standard statistical (Hardy-Weinberg) proportions. Structure can result from selection and mating among closely related individuals (e.g., siblings or cross-generational relatives) that are near one another. In studies of Pin us sylvestris (Scots pine) (Tigerstedt, 1 984) and Liriodendron tulipifera (tulip tree) (Brotschol et al., 1986), some evidence exists for such inbreeding or mating within localized populations. Thus, although the potential exists for both wide intermating and for neutral alleles to be randomly distributed, there is growing evidence that various genotypes in plant populations are not randomly distributed. Nonrandom mating within populations of continuous forest stands has been reported in several cases: Pinus ponderosa (western yellow or ponderosa pine) (Linhart et al., 1981a,b), Cryptomeria (red cedar) (Sake) and Park, 1971), Eucalyptus (eucalypts) (Moran and Hopper, 1987), and in some tropical species (O'Malley and Bawa, 1987~. A well-documented example of genetic structuring is a small population of Pinus tueda (loblolly pine). Allelic differences among groups have been reported in the continuous population and are assumed to be due to three separate regeneration episodes that established the current population (Roberds and Conkle, 1984~. Thus, genetic variations within and among popu- lations can be differentially distributed, and sampling or collecting that variation must include those structural variations and not depend on random methods. ESTIMATING GENETIC VARIATION Establishing priorities for the conservation, management, and use of tree genetic resources requires an understanding of the degree of diversity among and between specific populations. The scientific meth- ods used to distinguish levels of variation are the working tools for shaping decisions on which resources should be devoted to managing distinct tree characteristics. Two approaches have been used to study genetic diversity in tree populations: provenance testing and allozyme screening. The two approaches are used with different objectives in mind, and the results obtained from each have different applications in theory and practice. Provenance (geographic origin) testing is performed by gathering seeds from different populations and observing variation in performance (e.g., height, diameter, color, yield) in plants grown under uniform environmental conditions within one or more planting sites. The pop- ulations sampled are generally from different geographic and climatic regions. This approach constitutes the basis of provenance testing in forestry. Allozyme screening is based on the survey of genetic diversity

OCR for page 51
~ A A .. as revealed by variation in enzymes atspeciCc gene l!oci..~In general/ own provenance Sting and allozyme screening reve~althat, among pJ.a!n~ ~;orpi.ic~loci, and (3)~t~he Desalt number ofd~#eles~er~Iocus.l~ ~e~neral, ~resttrees am ~highIy h~ete~ozygou~s--the av<~ge~Ieve~l~f hete~Qzygosi~ ist~ice thatin~herbaceouspl~nts(Ha~mric~keti1.,1~979~.~milarkythep~ropo~ion of~olvnno~hicIodithose far enrich nu~erou~s~al~le)es~xist)Eisalso limb; ^ ~ most spewers show Gaelic Adaption at more tian.50 percent of the studiedly (T)bleS-2)s Al!houg) neck >chats> has bee~n~sirve!ed In a Abbe Anne of <~ecies~in~cl~ydin<-c~n~i.ter OCR for page 51
structure of Genetic Variation 161 torreyana and Abies bracteata have restricted ranges, but Pinus resinosa is widely distributed. Understanding the subtle factors that induce and maintain genetic diversity provides the scientific foundation for the optimal management of existing tree genetic variation. Estunates of Population Genetic Structures There is growing evidence that various genotypes in plant populations are not randomly distributed. Nonrandom distribution of genotypes in a population can make diverse sampling difficult. Selection and mating among related individuals (siblings or cross-generational relatives) in close proximity can affect genetic structure and thus sampling within the population. For several species, there are more homozygous individuals than would be expected from random mating, even within plots of 1 ha or less. This cannot be explained by selfing or other forms of inbreeding. In one such species, Pinus sylvestris, the excess of homozygotes could be due to very localized and specific male effectiveness (Tigerstedt, 1984~. This appears to be true for Liriodendron tulipifera (Brotschol, 1983) but only at the seedling stage, and it is not evident in the adult trees. Such effects appear to be even more pronounced in some tropical tree species, probably due to more limited pollen dispersal by animals (O'Malley and Bawa, 1987~. Pollen flow in these species appears to be very extensive, but the pollen distribution of individual trees may be highly localized. The net effect is that seed and seedling populations may be highly structured. The implication for sampling is that unless the sampling is spread over a wide area within the population, the collected seeds may represent a biased sample of genetic diversity within the population. Variation Among Populations Just as variation in species has implications for management, so does the distribution of that variation through the species. Forest tree species differ greatly with respect to genetic divergence among populations. There are two major problems in comparing levels of genetic differen- tiation within a species. First is the problem of scale; in some species, sampled populations are separated by several hundred kilometers, in others by only a few kilometers. Second, genetic divergence in some species has been studied on the basis of variation in morphological traits and in others by estimating variation in allozymes. The results obtained from the two approaches frequently do not agree. In many species, the evidence for genetic divergence is largely based on allelic diversity as revealed by allozyme studies. Several methods

OCR for page 51
floss: two mods Am 4,' :~s Was ~ +~ yews w ~ ~ :~. ~ ~ ~ sale ~ 7,~ k 91~ ~ ) Y IBM 'I ~5 ~ ^5 ~ I: ~ 5 15~*5 HAS *en; YE <;$ ~ YE *3:y 5 :~.'446~ If' = ~ 5 - S ides: ~5' ~5 If.... SIX: 'a: 'I; ~ .~,: s alms 1 ^~5 5 :~. ~ 'at 5 . - ~Apt x ~Gym ~HAY* ~Sly ~ t~( ~- ~=5 ~'~'.5 ~ ~'0< ~-' Am =5'Yak. S.~x s,, ^ ~ ~ ~ sew Is ~ ~ ~3 ~ ~,4 =< =) ~ ~ ~ I. .,*~< K- Hi _,--,,,' Is so I. ~ ..~. ~ s -I ., :=s ~Is - ~aim use s me Is ~Is em  =' Am semi,: ~seas 5 A MY ~Am* ~5-~54 ,s ~.~;s ', s ~ ~SA ~Y40: em ~,~S S,~ maxi 6; Em; ~ Y +5 Sol iI3~= ~= SS*S 5-~ ~ 0: ~{t- ~ 5 - 5.`' 5~ ~ I: ~ ~ ~ ~ (^ ~ ~ -* ~5 He, 5' ~ ~ ~ ~ ~ ~5 ~ ~ ~ .~ ~ ~5 ~ :5~. - *: Yen :* ~ I: ~O ~^: '< 'yip .^Pyr. 5 ~of* 5y'`~*~ y 'gyps ~He* ~ ~5~ you IS ~=~. ~ ~ ~5 ~ 3 ~5 - : =,. 54 W jS4,,j.* - * ~5 ~ ~ ~*I =5 at ~ =~5 y,40* ,l~* =5 S50~( ^~-S; ~ 5=S =. ~ >~5 =: 'A ~*:5' ~5 I.< IS* lo* 5-5 - 5 ~- 5. I: IS S'- ~ ~5 ~ 5,~j ~ 'So, ~ =5. , SSS~ cloy Saw. ~^ 'go' ~ 5 S-~ I ~ ~ ~ ye ,S - * ~ ~ IS:; HEY ~ 5 - S ~.t 5' ' - 5 =5 ~ ~ '= i> ='* US - - s Yip>*: ~ US'' ~ Son< - ' ' ~ ~ 5~5. IS :~4 >S Any :~, Soy} *= ~ SITS SUMS, =. ~ ~ -IS < 5 5.05 ~ 'em' S - S ~ ~ s s ,sS Is ~ - ; I've ~ BY; $ - > r'.,~,:> yt,,S S`'~y S - 5 =5 By - .5 it,,= :5~5, _' Son .Yr. Of i

OCR for page 51
63 _` us 0 0 0 {t ._ ~a, 0 0 0 ~ :, . ~, . _' 5 33 3 U 3 - e 3 A _ ~U ~o ~ ~ ~5 , e ~ ~ ~ ~ ~ Pa ~ on Z up ~ Z ~ ~ O ~ Z ~ ~0 ~ ~ ~ . . ~ ~ . ~u ~ ~ . u ~ . . ~ us . . . a) ~- v) ~ a.~ ~ ~ ~Q) en, (t) al (I us =,,~ 5- o =, ~ ~ O CL O c; o O ~ ~ ~ 0 ~Q O ~ CL O ~ Q _ ~ O O =, CL, O O ~ =, O Q U) U =, O O UO ~ O O ~ >, O =' O =, ~ C,) Q Q CO C~ Lr) ~ 00 ~C ~ ~ ~ ~3N ~ C~ ~ ~d~ ~u~ ~ oo ~ ~ O ~ ~ ~D Lr~ ~L~ ~ ~ IO ~O Lf) O ~ C~ o ~ oo ~ ~ ~ ~ ~ ~N ~ ~ O Lr) ~N C~ ~C~ ~ d~ CN ~ ~ ~ ~ CN L~ 1 ~ ~ O o o Lr) 1 C~ 1 ~ 1 0 C~) 1 oo ~ ~ oo ~D 1 o 1 ~ 1 ~- 1 ~) O O O 'S) 0 ~1 1 tN 1 ~O o o o ~o 1 ~ 1 ~o ~ ~ ~ o ~ ~o o cr ~ ~ ~ ~ ~ oo ~o ~ o ~ ~o o oo o o o oo ~ o o ~ C~ C~) ~) o ~ ~ oo o ~ ~ Lr) o C ~1 ~ ~ O O ~ ~ ~c ~ c~ c~ ~o ~ d~ ~Lr) ~O .. ...... . .... . .. .. . O O O O O O O O 0 00 0 0 0 0 0 O O O O O U) ~ $ C C C , v~ , v: C q) m ~ C ~ - ~ ~ ~ ~ ~ V) U) V) U) V) ~ ~ - C: ~: 0c 0c V) ~ C: ~ ~ ~ =: . . . . . _ <_) ~ ~ ~ ~ ~ ,~, ~; ~:

OCR for page 51
:~ a: - w~* ~- : ~sat ~ ~ ~= ~ , ~ :~ ~ ~ ~ :4 {- {~ ~ ~ = ~ ~ ~A~ By.: :~.~ -` :~ ' I:-, ,~ I, no. ~ ~ I' ~ ~ ^ - ~ ~ I =9_~~, ~ ~rho) ~.,, :.: ~ ~i: ~ it: it ) ~i. := 40 t ~SAW> t~~r ~ t~ ~ ~ ~ ~ ~ . - ~Am I; > - by' -= a_* rid ..~ ~ .- . ?~ 'ok., en. '<161~ If? _* By; . '~-,iS :~: 'my .,, - :~ :>~* an' i_, Cal ''by . - I ~ - .~: of go: :;Off ~By* ~4d I' ' ~ =' ~I' 2 ~t: 2- : ~:: :: ,- $ ` ~ - / ~ Id'; = ~J ~: i: ~. .. ~< if!= ~ -.,;-j ~ ,= ' - if: ,~. oaf = :: :, .,dj~,[ :~ ?.-,< ,, - `,, .. ~ ~ .'. :: ?~- :: :.: . ~I: ~t :: =\ ~'. <. ~i: I' ~-i ~NS ' i ~ ~ :'m' ~ - ) I' ~^ ~ - ~ (Not ~ ~ ~ ~ - : ~ At' ~ ~ ?~ ~ ~ - ~- mu. .~. Sac ram. ., ~ ~ ~ ~ ~ Y ~ ~ ~ ~ =' I- '= me,' ~ ~< .~ - .~ I. := . ~ I,,` ~? ,r=, ~ N.: ~- 7 ~t ~. ~4 - ~ ~ ,V O V - V w' ~ ~ V ~ ~ I' ~ ~ ~' ~ ~ ~q _ ~= W ~ ~/~

OCR for page 51
65 U) U) ~ CC o o o o sit 50 ~ ~ o o o o o o o o ~BUM o o o o ~ ~ 50 50 ~ 50 ~ . . . . U) ~ UC UC =4 Q AL O O O O ~== Cal ~ Cal 1 d ~Lr) O Lr) ~ ~ CO CO ~D ~D O ~ L~ . . . . O O O O ~: ~ ~: , ~ ~ ~ ~ C: ~ O ~ ~ ~ O C oo c~ C N ^` .. a, b4 aJ ~O on ~. _ ._ `~) ~.0 (t' I . O c) . O, ~_ ~ O .=~ ~ =; ~ ~ e5 D ,,; ~.t V O~ (= C~ cl5 ~ ~ ;: G ~ e ' _ C G tt ~ ~ _ _ _ O ~ ~ ~- , ~ ' _ ~ C o ~ ~ ~, ~ o C/ ~] 3 1 3 ~ ~ == 3 ~ o ~ CL ,: o ~ ~ ~ ~ ~ ~ A G ~ ~ oC '~ ~ ~ u a~ ~ ~ -. ~ ._ ._ g ~ ~ - O pL] =, =0, 3 6_ ~, ~ O U U] O ~ - ~- O ~

OCR for page 51
~ If ~ Knowledge about the d~i~sidbuti~on and reductive bi~ol~gv of tropical Mast trees is Mined throug~b botanical surveys and plant inv~n~es in the tonics. there a research assistant climbs a tree of the ~2 sp. to collect botanical stamens. Credit Calvin R. Sawing.

OCR for page 51
Structure of Genetic Variation / 67 exist not only for assessing the average levels of differentiation (Gre- gorius, 1984), but also for determining which loci or which populations are most deviant (Gregorius and Roberds, 1986~. Among these, Nei's measures are most commonly reported (Nei, 1972~. It can be shown that species vary considerably in terms of both the patterns and the levels of genetic variability among populations. Within most species, some hierarchy of subdivisions exists due to interpopu- lational segregation within localized plots and to individual tree inbreed- ing, as in Abies fraseri (Ross, 1988~. Typically, widespread species, such as Pseudotsuga menziesii (Douglas fir), show geographical or ecotypic differentiation as, for example, between coastal and inland populations. Moreover, within each geographical area, large interpopulational dif- ferences may exist. Such species may thus show high variability within and among populations. Other species, such as Pinus longaeva (Great Basin bristlecone pine), show high variability within populations but low differentiation among populations (Hiebert and Hamrick, 1983~. In species such as Pinus resinosa and Abies bracteata, the level of variation is low both within and among populations. By contrast, the two populations of Pinus torreyana, each of which is genetically uniform, are quite distinct from each other. The level of genetic differentiation may also vary within a species. Pinus monticola (western white pine), for example, is highly variable in California, but not in other parts of its range (Steinhoff et al., 1983~. A very preliminary analysis of a few tropical forest tree species reveals that levels of population genetic divergence are similar to those observed for temperate tree species. This is contrary to expectations based on the limited gene dispersal that might be typical of many tropical species whose populations are more isolated (Bawa, 1976~. The level of genetic differentiation among populations has been quantified in many species. In most species, particularly the widespread ones, 2 to 16 percent of the total genetic variability is due to interpop- ulation differences (Fins and Seeb, 1986; Moran and Hopper, 1987~. Mating systems and the geographical range of species seem to have significant effects on the level of genetic variation among populations (Hamrick, 1983~. Self-pollinating species show more interpopulational divergence than outcrossing species (Hamrick, 1983~. Regional and localized species also exhibit greater differentiation among populations than widespread species (Moran and Hopper, 1987~. The conservation implications of the available data on genetic variation are that, in the case of outcrossing and widespread species, a few populations in each geographical zone may conserve much of the genetic

OCR for page 51
68 / Forest Trees diversity. For inbred and regional or localized species, however, many populations may be required to conserve a significant proportion of genetic variation (see Chapter 4~. Allozyme Studies and Morphological Variation It is not known whether genetic differentiation inferred from allozyme studies is neutral, as generally assumed, or if it is indicative of a selective advantage or disadvantage. This is important because if much of the variation seen in allozymes were to be devoid of evolutionary significance (neutral), it would be difficult to justify maintaining multiple populations of trees to preserve their potentially useful genetic diversity based solely on allozyme variations. Unfortunately, allozyme studies in forest trees generally have not been accompanied by parallel work on variation in morphological traits that are known to contribute to the fitness of plants. In some studies, agreement between variation in allozymes and variation in other quantitative traits has been reported (Hamrick, 1983~. In several species, however, the level of genetic divergence revealed by variation in allozymes is not correlated with morphological variability detected by provenance research (Libby and Critchfield, 1987; Moran and Adams, 1989; Namkoong and Kang, 1990~. In particular, large-scale genetic variation along a geographic gradient (clinal variation) in Picea sitchensis (Sitka pine) or the extensive variety of ecotypes in Pseudotsuga menziesii is not revealed by allozyme studies (Falkenhagen, 1985~. Decades of provenance research have shown that most forest tree species exhibit considerable population divergence in genetically based traits of direct survival value. There remains, however, a need to conduct parallel allozyme and morphometric studies to determine to what degree data on isoenzymes can be used as indicators of morphological variation, or to monitor changes in gene freauencv due to environmental or ecological events. Allele Frequencies, Gene Flow, and Selection The importance of interpopulational differentiation lies not only in whether alleles would be missed by sampling only a few populations, but in the occurrence of different combinations of alleles and their frequencies in different populations. This could make certain alleles or combinations of alleles easier to obtain from some populations than others. For outcrossing species with both large population sizes and high migration rates, the existence of alleles unique to a population is unlikely. In such species, alleles are likely to be widely distributed

OCR for page 51
Afro ~ ~3 doff / ~ Lon~te!~ ec~o{cal studies on the I of Spies floss 1~ It Crests am piano Inducted so that resews of adequate size may ~ es6bli~s~hed to pe~etusp!te spades dived. ~ i~o~in~t~ plant of Other Brazilian Insti~tuto Nan. de ~sq~s~s da A~~n~ (~lN~PA) <~( Ace U.S. World Wildly Bound is s~$ud~yi~ reserves Unsung from I ha {pictured abode) ~ l~O~,Oqp ba in -3~1 Bath ~~ are inflated by c}ead~ adjacent land. Conduit: Douglas Fly. -~ith~o~.t much vadatBon~i:n average / Enemies. Nat~ralselectio~n' h!oY'~- eve~r/c~n p~yastrong~rolein.~ausing di~ere~ndaC~o~na!mosn~ pop~ulahons, especially~i~ffiesel~ecionpressL~resa~ height mig~ationisl~ob orepisodic~, I_ . ~ ~ The~potenba~l~r~ideseedorpo~llend~ispersala~>ctsbened~cstru~cture. for some species, even those as v~i.d~espread as Fires fords Sian the southeastern United States,this may not always betrue. Reproduction ca.n,in some ~ me periods, be episodic in relatively so ~] p~tcbe~a~nd thuslimit the populad.o~n sa m pig in each pat~ch.ll is ~ oul].ge~erste a g~ne~ca.~nd d!ennograph.~c~osaic.S~ch naosaicsca~n disappearif~:ide- scaleclea~ngoccu~Roberdsa~nd C~onkle,1984~.Ceno~p~icdist~b~tio~n of any reestablished population then depends on the mating pool Taunt within the limited time pedod of forest ~re~n~e=~on. For species witch mom ~std~cted pollen or weds migration, mosaics Kitchen Faust sands are expressed as Awe inte~opulabonal vadabons of Ionizer dungeon. Selection awes operating on diffedng gene f~q~ue~ncies can abut the d~bibuLon oigenedcvadabon among and Within those popu~bons.

OCR for page 51
70 / Forest Trees The maintenance of the heterogeneity of allele frequencies that results from selective differences in the absence of mating barriers is possible but usually very difficult. Thus, some restrictions on gene flow are usually necessary for populations to differentiate under natural condi- tions. Over large distances of dense stands, the area of interbreeding neighborhoods can be small for some tree species (Wright, 1962~. In fact, most tree species do display some degree of interpopulational differentiation in at least some phenotypic traits. For breeding purposes, it may be of direct importance to capture and make use of such interpopulational differences. The genetic structure of populations thus provides a good average view of the amount of subdivision among population samples. Mating structures can be complex, and no single statistic can explain all of the important features of allelic distributions. Often differences between pollen and seed migration rates exist, and the extent to which they coincide with selective differences affects the distribution of alleles, allelic extinction rates (rate of gene loss), and the probability distributions of unique alleles (Gregorius and Namkoong, 1983; Namkoong and Gregorius, 19851. Differences also exist in the patterns of population subdivision indi- cated by isozyme versus morphological traits, in the traits associated with broad climatic adaptability (e.g., bud break or bud set times), and in the traits associated with local site differences, such as elevation. The boundaries of selection do not act consistently among traits or among loci. There is often reason to suspect that local, naturally occurring populations are neither optimally reproductively fit (Eriksson et al., 1972) nor ideal for selection in production forestry breeding, even within their area of origin (Namkoong, 1969~. Natural occurrence of a species in a region, therefore, should not be the sole criterion for its development as a production crop. Population Complexity Populations of trees evolve from different genetic structures and with the confounding effects of inbreeding and selection. A major analytical difficulty is to discern which factors are most important in generating the current and evolving patterns of genetic distribution in the popu- lation. Given the complexity of the mating systems of many tree species and the ecological instability of stand boundaries and gap distributions over time, great potential exists for many locally differentiated popu- lations to have existed in the past or to be generated in the future. This

OCR for page 51
Structure of Genetic Variation 171 can significantly complicate sampling the current variation. It must also be considered when managing a sampled set of genotypes or conser- vation stands. CONCLUSIONS Tree species are most commonly and effectively managed in planted stands. In many instances, trees will be maintained in or near their natural habitat and be classified as in situ. They can, for a number of objectives, be grown and sustained in new environments ex situ. For both types of management, it is of fundamental importance to under- stand the biological factors that influence the structure of genetic variation within tree populations. The current state of knowledge about breeding systems is sparse, as is the knowledge of the structure of populations, especially in tropical species. In nature, trees exhibit a high degree of heterogeneity (genetic differences), and the goal of conser- vation and management activities must be to capture and maintain that nonuniform genetic configuration. The criteria to be used for sampling trees to be conserved and maintained will depend on the breeding system of each species. If knowledge of the reproductive biology of a species is incomplete, collection activities should be conducted at all extremes of population occurrence and include a high level of redundancy, even in the central areas of occurrence. The designation of large areas for in situ conservation is especially critical in tropical regions where tree species occur in lower densities as part of complex ecosystems that are biologically poorly understood. RECOMMENDATIONS To support conservation efforts, study of the patterns of genetic variation in tree populations should be accelerated and expanded in scope, especially in the tropics and subtropics. Although genetic variation has been surveyed in a wide range of species, the existing data base is largely derived from the study of conifers in the north temperate zone. Similar data for tropical regions are lacking. A very preliminary analysis of a few tropical forest tree species reveals that levels of genetic divergence within populations are similar to those observed for temperate tree species. This is contrary to what the very different reproductive mechanisms of many tropical species have suggested. Thus, further study is needed before this information can be generally applied to conserving tropical tree species.

OCR for page 51
72 / Forest Trees Research is needed to elucidate the distribution and structure of genetic variation-especially for tropical trees and to support conservation efforts. Study of the patterns of genetic variation in tree populations should be accelerated and expanded in scope, especially in the tropics and subtropics. Genetic management and conservation programs should be applied to many more species and populations, particularly those that are endangered and those that are potentially useful. Inventories of genetic resources and patterns of variation should be accelerated and expanded to more areas, especially in the tropics. Genetic management techniques for resource areas that are managed for agroforestry and for industrial forestry should be studied to develop inexpensive means for monitoring the distribution of genetic variation and ensuring the main- tenance of genetic diversity. Research on population sizes, structures, dynamics, and reproductive systems should also be expanded, particularly in the following three areas: Research should be increased on reproductive systems, the genetic architecture of populations, and the minimum viable population size of trees. Such information is essential for the design and management of in situ and ex situ conservation stands. Knowledge must be gained of key interactions in complex tropical communities to support management of in situ stands and to prevent large-scale changes in population structures and species composition. Efforts should be made to anticipate the effect of global climatic change and pollution on the geographical distribution of species and on in situ reserves. .