Overview - PowerPoint PPT Presentation

About This Presentation
Title:

Overview

Description:

Overview Armillaria bulbosa (gallica) Known as the Humungous Fungus, or honey mushroom Form rhizomorphs, which make up much of the humungous part – PowerPoint PPT presentation

Number of Views:183
Avg rating:3.0/5.0
Slides: 134
Provided by: WillV152
Category:

less

Transcript and Presenter's Notes

Title: Overview


1
Overview
  • Armillaria bulbosa (gallica)
  • Known as the Humungous Fungus, or honey mushroom
  • Form rhizomorphs, which make up much of the
    humungous part
  • Basidiocarp cap 6 cm in diameter, stem is 5-10
    cm tall
  • Facultative tree root pathogen

2
Life cycle Reproduction
  • Sexual
  • Basidiocarps release spores (n) after karyogamy
    and meiosis
  • 2 mating-type loci, each with multiple alleles in
    the population
  • Isolates (n) must have different alleles at two
    mating type loci to be sexually compatible
  • Asexual
  • vegetative spreading of rhizomorph
  • The large mass of rhizomorph that is genetically
    isolated is called a clone

3
Building up the question
  • By extending the areas sampled in subsequent
    years, we were finally able to delimit the large
    area occupied by this genotype and then go on to
    show that this genotype likely represents and
    individual
  • - Myron Smith

4
Researchers Question
  • The clonal individual is especially difficult
    to define because the network of hyphae is
    underground
  • How do you unambiguously identify an individual
    fungi within a local population?

5
Approach
  • 1. Collect samples
  • 2. Check mating type
  • - Somatic compatibility test
  • - Distrubution of mating-type alleles
  • 3. Molecular testing
  • - RFLP
  • - RAPD
  • 4. Statistics
  • 5. More testing

6
Methods and Materials 1
  • 1. Collecting samples
  • Researcher collected samples over a 30 hectare
    area by baiting Armillaria with poplar stakes and
    taking tissues and spores
  • They then grew the successfully colonized stakes
    in soil taken from the study site
  • Each fungal colony cultured was called an isolate.

7
Methods and Materials 2
Example (not Armillaria)
  • 2. Checking mating type
  • - Somatic incompatibilityFor two fungal
    isolates to fuse, all somatic compatibility loci
    must be the same.
  • Fusion means theyre clones ?

8
Methods and Materials 2
  • 2. Checking mating type
  • - Distrubution of mating alleles
  • Mating occurs only when coupled isolates have
    different alleles at two unlinked, multiallelic
    loci A and B. (They have an incompatibility
    system)
  • If fruit bodies had the same alleles at A and B,
    and were collected from the same area, they were
    assumed to be from the same clone

9
Result 1
  • Somatic compatilbilty
  • isolates from vegetative mycelium from a large
    sampling area fused
  • Mating alleles
  • They had the same mating type

10
Result 1
  • Clone 1 was found to exceed 500 m in diameter
  • Used previously collected mtDNA restriction
    fragment patterns

11
Sensitivity of Approach
  • Problem These tests alone are not enough to
    distinguish a clone from closely related
    individuals

12
Why?
  • Q The first two tests were not sensitive enough
    to tell a clone from a close relativeWhy?
  • A Spores from same point source have the same
    mating-type alleles, but the offspring they
    produce after inbreeding are genetically
    distinct.

13
Methods and Materials 3
  • 3. Molecular Testing
  • - RFLP analysis at 5 polymorphic, heterozyg. loci
    of mtDNA from Clone 1
  • - RAPD analysis at 11 loci

14
RAPDS vs. RFLPs
  • Use 1 short PCR primer
  • When it finds match on template at a distance
    that can be amplified (primer binds twice within
    50 to 2000 bp) RAPD amplicon
  • Dominant, annoymous
  • Total genomic, vs single locus
  • Use endonuclease to digest DNA at specific
    restriction site
  • Run digest and see how amplicon was cut
  • Single locus is co-dominant

15
Result 2
  • RFLP
  • All 5 loci from Clone 1 were heterozygous and
    identical (both alleles present at loci
    1,1)
  • RAPD
  • All 11 RAPD products were present in all
    vegetative isolates

16
Statistical Analysis
  • The probability of retaining heterozygosity at
    each parental locus in an individual produced by
    mating of sibling monospore isolates
  • 0.0013
  • So they were pretty confident that cloning was
    responsible for their results, not inbreeding

17
More testing, just in case
  • To be completely confident, they tested
  • 1) that nearby Clone 2 was different and lacked 5
    of the Clone 1 heterozyg. RAPD fragments,
  • 2) more loci, totaling
  • 20 RAPD fragments
  • 27 nuclear DNA RFLP fragments
  • all were identical in Clone 1

18
Sensitivity of RAPDs
  • Tested on subset of spores from same basidiocarp
  • RAPDs differentiated among full sibs

19
Conclusions
  • Somatic compatibility, mating allele loci, mtDNA,
    RFLP, and RAPD tests all indicate that a single
    organism could indeed occupy a 15 hectare area

20
Conclusions
  • The larger individual, Clone 1 was estimated to
    weigh 9700 kg and be over 1500 years old

21
Implications
  • ?????
  • Fungi are one of the oldest and largest organisms
    on the planet
  • Recycle nutrientsvery important!
  • Armillaria bulbosa also a pathogen its effects
    on forest above may be huge as well.

22
HOST-SPECIFICITY
  • Biological species
  • Reproductively isolated
  • Measurable differential size of structures
  • Gene-for-gene defense model
  • Sympatric speciation Heterobasidion, Armillaria,
    Sphaeropsis, Phellinus, Fusarium forma speciales

23
(No Transcript)
24
Phylogenetic relationships within the
Heterobasidion complex
Fir-Spruce
Pine Europe
Pine N.Am.
25
The biology of the organism drives an epidemic
  • Autoinfection vs. alloinfection
  • Primary spreadby spores
  • Secondary spreadvegetative, clonal spread, same
    genotype . Completely different scales (from
    small to gigantic)
  • Coriolus
  • Heterobasidion
  • Armillaria
  • Phellinus

26
(No Transcript)
27
OUR ABILITY TO
  • Differentiate among different individuals
    (genotypes)
  • Determine gene flow among different areas
  • Determine allelic distribution in an area

28
WILL ALLOW US TO DETERMINE
  • How often primary infection occurs or is disease
    mostly chronic
  • How far can the pathogen move on its own
  • Is the organism reproducing sexually? is the
    source of infection local or does it need input
    from the outside

29
IN ORDER TO UNDERSTAND PATTERNS OF INFECTION
  • If John gave directly Mary an infection, and Mary
    gave it to Tom, they should all have the same
    strain, or GENOTYPE (comparisonsecondary spread
    among forest trees)
  • If the pathogen is airborne and sexually
    reproducing, Mary John and Tom will be infected
    by different genotypes. But if the source is the
    same, the genotypes will be sibs, thus related

30
Recognition of self vs. non self
  • Intersterility genes maintain species gene pool.
    Homogenic system
  • Mating genes recognition of other to allow for
    recombination. Heterogenic system
  • Somatic compatibility protection of the
    individual.

31
Recognition of self vs. non self
  • What are the chances two different individuals
    will have the same set of VC alleles?
  • Probability calculation (multiply frequency of
    each allele)
  • More powerful the larger the number of loci
  • and the larger the number of alleles per locus

32
Recognition of self vs. non self
  • It is possible to have different genotypes with
    the same vc alleles
  • VC grouping and genotyping is not the same
  • It allows for genotyping without genetic tests
  • Reasons behing VC system protection of
    resources/avoidance of viral contagion

33
Somatic incompatibility
34
More on somatic compatibility
  • Perform calculation on power of approach
  • Temporary compatibility allows for cytoplasmic
    contact that then is interrupted this temporary
    contact may be enough for viral contagion

35
SOMATIC COMPATIBILITY
  • Fungi are territorial for two reasons
  • Selfish
  • Do not want to become infected
  • If haploids it is a benefit to mate with other,
    but then the nn wants to keep all other
    genotypes out
  • Only if all alleles are the same there will be
    fusion of hyphae
  • If most alleles are the same, but not all, fusion
    only temporary

36
SOMATIC COMPATIBILITY
  • SC can be used to identify genotypes
  • SC is regulated by multiple loci
  • Individual that are compatible (recognize one
    another as self, are within the same SC group)
  • SC group is used as a proxy for genotype, but in
    reality, you may have some different genotypes
    that by chance fall in the same SC group
  • Happens often among sibs, but can happen by
    chance too among unrelated individuals

37
Recognition of self vs. non self
  • What are the chances two different individuals
    will have the same set of VC alleles?
  • Probability calculation (multiply frequency of
    each allele)
  • More powerful the larger the number of loci
  • and the larger the number of alleles per locus

38
Recognition of self vs. non selfprobability of
identity (PID)
  • 4 loci
  • 3 biallelelic
  • 1 penta-allelic
  • P 0.5x0.5x0.5x0.20.025
  • In humans 99.9, 1000, 1 in one million

39
INTERSTERILITY
  • If a species has arisen, it must have some
    adaptive advantages that should not be watered
    down by mixing with other species
  • Will allow mating to happen only if individuals
    recognized as belonging to the same species
  • Plus alleles at one of 5 loci (S P V1 V2 V3)

40
INTERSTERILITY
  • Basis for speciation
  • These alleles are selected for more strongly in
    sympatry
  • You can have different species in allopatry that
    have not been selected for different IS alleles

41
MATING
  • Two haploids need to fuse to form nn
  • Sex needs to increase diversity need different
    alleles for mating to occur
  • Selection for equal representation of many
    different mating alleles

42
MATING
  • If one individuals is source of inoculum, then
    the same 2 mating alleles will be found in local
    population
  • If inoculum is of broad provenance then multiple
    mating alleles should be found

43
MATING
  • How do you test for mating?
  • Place two homokaryons in same plate and check for
    formation of dikaryon (microscopic clamp
    connections at septa)

44
Clamp connections
45
MATING ALLELES
  • All heterokaryons will have two mating allelels,
    for instance a, b
  • There is an advantage in having more mating
    alleles (easier mating, higher chances of finding
    a mate)
  • Mating allele that is rare, may be of migrant
    just arrived
  • If a parent is important source, genotypes should
    all be of one or two mating types

46
Two scenarios
  • A, A, B, C, D, D, E, H, I, L
  • A, A, A,B, B, A, A

47
Two scenarios
  • A, A, B, C, D, D, E, H, I, L
  • Multiple source of infections (at least 4
    genotypes)
  • A, A, A,B, B, A, A
  • Siblings as source of infection (1 genotype)

48
SEX
  • Ability to recombine and adapt
  • Definition of population and metapopulation
  • Different evolutionary model
  • Why sex? Clonal reproductive approach can be very
    effective among pathogens

49
Long branches in between groups suggests no sex
is occurring in between groups
Fir-Spruce
Pine Europe
Pine N.Am.
50
Small branches within a clade indicate sexual
reproduction is ongoing within that group of
individuals
NA S
NA P
EU S
890 bp CIgt0.9
EU F
51
Index of association
  • Ia if same alleles are associated too much as
    opposed to random, it means sex is not occurring
  • Association among alleles calculated and compared
    to simulated random distribution

52
Evolution and Population genetics
  • Positively selected genes
  • Negatively selected genes
  • Neutral genes normally population genetics
    demands loci used are neutral
  • Loci under balancing selection..

53
Evolution and Population genetics
  • Positively selected genes
  • Negatively selected genes
  • Neutral genes normally population genetics
    demands loci used are neutral
  • Loci under balancing selection..

54
Evolutionary history
  • Darwininan vertical evolutionary models
  • Horizontal, reticulated models..

55
Phylogenetic relationships within the
Heterobasidion complex
Fir-Spruce
Pine Europe
Pine N.Am.
56
Geneaology of S DNA insertion into P ISG
confirms horizontal transfer.Time of
cross-over uncertain
NA S
NA P
EU S
890 bp CIgt0.9
EU F
57
Because of complications such as
  • Reticulation
  • Gene homogeneization(Gene duplication)
  • Need to make inferences based on multiple genes
  • Multilocus analysis also makes it possible to
    differentiate between sex and lack of sex
    (Iaindex of association), and to identify
    genotypes, and to study gene flow

58
Basic definitions again
  • Locus
  • Allele
  • Dominant vs. codominant marker
  • RAPDS
  • AFLPs

59
How to get multiple loci?
  • Random genomic markers
  • RAPDS
  • Total genome RFLPS (mostly dominant)
  • AFLPS
  • Microsatellites
  • SNPs
  • Multiple specific loci
  • SSCP
  • RFLP
  • Sequence information
  • Watch out for linked alleles (basically you are
    looking at the same thing!)

60
Example 2 Distinguishing taxa in the Pleurotus
eryngii (King Oyster Mushroom) complex using
AFLPs (Urbanelli et al., Appl. Microbiol.
Biotechnol. (2007) 74592-600)
Photo The New York Times
Photo Wikimedia Commons
60
61
Distinguishing taxa in the Pleurotus eryngii
(King Oyster Mushroom) complex using AFLPs
  • Goal to determine, using multilocus genotypes,
    whether the distinction between Pleurotus
    eryngii, P. ferulae, and P. eryngii var.
    nebrodensis is supported by genetic data
  • 90 populations sampled
  • 94 AFLP loci scored

61
P. ferulae (http//steinpilz.up.seesaa.net)
62
Sample AFLP Gel
62
63
AFLP Data Map from Urbanelli et al. (2007)
63
64
AFLP Data Map with UPGMA dendogram from Urbanelli
et al. (2007)
64
65
RAPDS use short primers but not too short
  • Need to scan the genome
  • Need to be readable
  • 10mers do the job (unfortunately annealing
    temperature is pretty low and a lot of priming
    errors cause variability in data)

66
RAPDS use short primers but not too short
  • Need to scan the genome
  • Need to be readable
  • 10mers do the job (unfortunately annealing
    temperature is pretty low and a lot of priming
    errors cause variability in data)

67
RAPDS can also be obtained with Arbitrary Primed
PCR
  • Use longer primers
  • Use less stringent annealing conditions
  • Less variability in results

68
Result series of bands that are present or
absent (1/0)
69
(No Transcript)
70
Root disease center in true fir caused by H.
annosum
71
Ponderosa pine
Incense cedar
72
Yosemite Lodge 1975 Root disease centers
outlined
73
Yosemite Lodge 1997 Root disease centers
outlined
74
WORK ON PINES HAD DEMONSTRATED INFECTIONS ARE
MOSTLY ON STUMPS
  • Use meticulous field work and genetics
    information to reconstruct disease from infection
    to explosion
  • On firs/sequoia if the stump theory were also
    correct we would find a stump within the outline
    of each genotype

75
(No Transcript)
76
(No Transcript)
77
(No Transcript)
78
(No Transcript)
79
(No Transcript)
80
(No Transcript)
81
Are my haplotypes sensitive enough?
  • To validate power of tool used, one needs to be
    able to differentiate among closely related
    individual
  • Generate progeny
  • Make sure each meiospore has different haplotype
  • Calculate P

82
RAPD combination1 2
  • 1010101010
  • 1010101010
  • 1010101010
  • 1010101010
  • 1010000000
  • 1011101010
  • 1010111010
  • 1010001010
  • 1011001010
  • 1011110101

83
Conclusions
  • Only one RAPD combo is sensitive enough to
    differentiate 4 half-sibs (in white)
  • Mendelian inheritance?
  • By analysis of all haplotypes it is apparent that
    two markers are always cosegregating, one of the
    two should be removed

84
If we have codominant markers how many do I need
  • IDENTITY tests probability calculation based
    on allele frequency Multiplication of
    frequencies of alleles
  • 10 alleles at locus 1 P10.1
  • 5 alleles at locus 2 P20,2
  • Total P P1P20.02

85
Have we sampled enough?
  • Resampling approaches
  • Saturation curves
  • A total of 30 polymorphic alleles
  • Our sample is either 10 or 20
  • Calculate whether each new sample is
    characterized by new alleles

86
Saturation (rarefaction) curves
No Of New alleles
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20
87
Dealing with dominant anonymous multilocus markers
  • Need to use large numbers (linkage)
  • Repeatability
  • Graph distribution of distances
  • Calculate distance using Jaccards similarity
    index

88
Jaccards
  • Only 1-1 and 1-0 count, 0-0 do not count
  • 1010011
  • 1001011
  • 1001000

89
Jaccards
  • Only 1-1 and 1-0 count, 0-0 do not count
  • A 1010011 AB 0.6 0.4 (1-AB)
  • B 1001011 BC0.5 0.5
  • C 1001000 AC0.2 0.8

90
Now that we have distances.
  • Plot their distribution (clonal vs. sexual)

91
Now that we have distances.
  • Plot their distribution (clonal vs. sexual)
  • Analysis
  • Similarity (cluster analysis) a variety of
    algorithms. Most common are NJ and UPGMA

92
Now that we have distances.
  • Plot their distribution (clonal vs. sexual)
  • Analysis
  • Similarity (cluster analysis) a variety of
    algorithms. Most common are NJ and UPGMA
  • AMOVA requires a priori grouping

93
AMOVA groupings
  • Individual
  • Population
  • Region
  • AMOVA partitions molecular variance amongst a
    priori defined groupings

94
Example
  • SPECIES X 50blue, 50 yellow

95
AMOVA example
Scenario 1
Scenario 2
v
POP 1
POP 2
v
96
Expectations for fungi
  • Sexually reproducing fungi characterized by high
    percentage of variance explained by individual
    populations
  • Amount of variance between populations and
    regions will depend on ability of organism to
    move, availability of host, and
  • NOTE if genotypes are not sensitive enough so
    you are calling the same things that are
    different you may get unreliable results like 100
    variance within pops, none among pops

97
Results Jaccard similarity coefficients
P. nemorosa
P. pseudosyringae U.S. and E.U.
98
P. pseudosyringae genetic similarity patterns are
different in U.S. and E.U.
99
Results P. nemorosa
100
(No Transcript)
101
The scale of disease
  • Dispersal gradients dependent on propagule size,
    resilience, ability to dessicate, NOTE not
    linear
  • Important interaction with environment, habitat,
    and niche availability. Examples Heterobasidion
    in Western Alps, Matsutake mushrooms that offer
    example of habitat tracking
  • Scale of dispersal (implicitely correlated to
    metapopulation structure)---

102
(No Transcript)
103
(No Transcript)
104
(No Transcript)
105
RAPDSgt not used often now
106
RAPD DATA W/O COSEGREGATING MARKERS
107
(No Transcript)
108
PCA
109
AFLP
  • Amplified Fragment Length Polymorphisms
  • Dominant marker
  • Scans the entire genome like RAPDs
  • More reliable because it uses longer PCR primers
    less likely to mismatch
  • Priming sites are a construct of the sequence in
    the organism and a piece of synthesized DNA

110
How are AFLPs generated?
  • AGGTCGCTAAAATTTT (restriction site in red)
  • AGGTCG CTAAATTT
  • Synthetic DNA piece ligated
  • NNNNNNNNNNNNNNCTAAATTTTT
  • Created a new PCR priming site
  • NNNNNNNNNNNNNNCTAAATTTTT
  • Every time two PCR priming sitea are within
    400-1600 bp you obtain amplification

111
(No Transcript)
112
(No Transcript)
113
Distances between study sites
White mangroves Corioloposis caperata
114
Forest fragmentation can lead to loss of gene
flow among previously contiguous populations.
The negative repercussions of such genetic
isolation should most severely affect highly
specialized organisms such as some
plant-parasitic fungi.
AFLP study on single spores
Coriolopsis caperata on Laguncularia racemosa
115
(No Transcript)
116
(No Transcript)
117
Using DNA sequences
  • Obtain sequence
  • Align sequences, number of parsimony informative
    sites
  • Gap handling
  • Picking sequences (order)
  • Analyze sequences (similarity/parsimony/exhaustive
    /bayesian
  • Analyze output CI, HI Bootstrap/decay indices

118
Using DNA sequences
  • Testing alternative trees kashino hasegawa
  • Molecular clock
  • Outgroup
  • Spatial correlation (Mantel)
  • Networks and coalescence approaches

119
(No Transcript)
120
(No Transcript)
121
From Garbelotto and Chapela, Evolution and
biogeography of matsutakes
Biodiversity within species as significant as
between species
122
Microsatellites or SSRs
  • AGTTTCATGCGTAGGT CG CG CG CG CG
    AAAATTTTAGGTAAATTT
  • Number of CG is variable
  • Design primers on FLANKING region, amplify DNA
  • Electrophoresis on gel, or capillary
  • Size the allele (different by one or more
    repeats if number does not match there may be
    polimorphisms in flanking region)
  • Stepwise mutational process (2 to 3 to 4 to 3 to2
    repeats)

123
ACACACACACACACACAC
  • MS18 (AC)38 218 bp
  • (AC)39 220 bp
  • (AC)40 222 bp
  • MS43a (CAGA)70 373 bp
  • MS43a (CAGA)71 377 bp
  • MS43a (CAGA)72 381 bp
  • (220-218)2 22
  • (222-218)2 42
  • (377-373)2 42
  • (381-373)2 82
  • (39-38)2 12
  • (40-38)2 22
  • (71-70)2 12
  • (72-70)2 22

AMOVA Analysis of Molecular Variance
123
124
Example 1 Origins of the Sudden Oak Death
Epidemic in California(Mascheretti et al.,
Molecular Ecology (2008) 17 2755-2768)
Photo UC Davis
Photo www.membranetransport.org
124
Photo Northeast Plant Diagnostic Network
125
Spatial autocorrelation
Within approx. 100 meters the genetic structure
correlates with the geographical distance
10 100 1000
Geographical distance (m)
125
126
Spatial autocorrelation
Morans I (coefficient of departure from spatial
randomness) correlates with distance up to
Distribution of genotypes (6 microsatellite
markers) in different populations of P.ramorum
in California
126
127
NJ tree of P. ramorum populations in California
HU-1
MA-1
HU-2
MA-2
SC-2 MO-1 MO-2
127
128
Example microsatellites genotyping of P.
ramorum isolates
  • Phytophthora ramorum (Oomycete)
  • causal agent of Sudden Oak Death (SOD) first
    reported in California in 1994
  • SOD affects tanoak (Lithocarpus densiflora),
    coast live oak (Quercus agrifolia), Californian
    black oak (Quercus kelloggii), and Canyon live
    oak (Quercus chrysolepis)
  • P.ramorum also cause a disease characterized
    mostly by leaf blight and/or branch dieback in
    over 100 species of both wild and ornamental
    plants, including California bay laurel
    (Umbellularia cailfornica), California redwood
    (Sequoia sempervirens), Camellia and
    Rhododrendron species

Collection of infected bay leaves from several
forests in Sonoma, Monterey, Marin, Napa,
Alameda, San Mateo
128
129
Microsatellites (I)mating type A1 (EU) and
mating type A2 (US)
  • A2 (US) A1 (EU)
  • Locus 29 325/ - 325/337
  • -/337
  • Locus 33 315/337 325/337
  • Locus 65 234/252 236/244
  • 220/222

129
130
Microsatellites (II)mating type A2 (US)
  • MS39a (GA)11 129 bp
  • MS39b (GA)4(GATA)31 234, 242, 246, 250, 254 bp
  • MS43a (CAGA)69 309, 329, 349, 353, 357, 361,
    365, 369, 373, 377, 381 bp
  • MS43b (CAGA)75()(CAGA)16 416, 420, 466, 470,
    474, 476, 478, 482, 486, 490, 494, 498 bp
  • MS45 (TCCG)11 167, 183, 187 bp
  • MS18 (AC)39 218, 220, 222, 254, 264, 272, 274,
    276, 278, 282 bp
  • MS64 (CT)16 342, 374, 376, 378 bp

130
131
  • Ind. MS39a MS39b MS43a MS43b MS45 MS18 MS64 Matin
    g type
  • 1 129-129 246-246 369-369 486-486 167-187 220-278
    342-374 A1
  • 2 129-129 246-246 369-369 486-486 167-187 220-278
    342-374 A1
  • 3 129-129 246-246 373-373 486-486 167-187 220-274
    342-374 A1
  • 4 129-129 246-246 373-373 486-486 167-187 220-27
    4 342-378 A1
  • 5 129-129 246-246 373-373 486-486 167-187 220-27
    4 342-378 A1
  • 6 129-129 246-246 373-373 486-486 167-187 220-27
    4 342-378 A1
  • 7 129-129 246-246 373-373 486-486 167-187 220-27
    8 342-378 A1
  • 8 129-129 246-246 373-373 486-486 167-187 220-27
    8 342-374 A1
  • 9 129-129 250-250 369-369 486-486 167-187 220-278
    342-374 A1
  • 10 129-129 250-250 369-369 486-486 167-187 220-2
    78 342-374 A1
  • 11 129-129 250-250 369-369 486-486 167-187 220-27
    8 342-374 A1
  • 12 129-129 250-250 377-377 490-490 167-187 220-2
    78 342-374 A1
  • 13 129-129 250-250 377-377 490-490
    167-187 220-278 342-381 A1
  • 14 129-129 250-250 377-377 490-490
    167-187 220-278 342-381 A1
  • 15 129-129 250-250 377-377 490-490
    167-187 220-278 342-381 A1
  • 16 129-129 246-246 377-377 490-490
    167-187 220-278 342-374 A1
  • 17 129-129 246-246 377-377 486-486 167-187 220-2
    78 342-374 A1

131
132
Ind. MS39a MS39b MS43a MS43b MS45 MS18 MS64 Matin
g type 1 11-11 32-32 69-69 17-17 11-16 39-68 18-
29 A1 2 11-11 32-32 69-69 17-17 11-16 39-68
18-29 A1 3 11-11 32-32 70-70 17-17 11-16
39-66 18-29 A1 4 11-11 32-32 70-70 17-17
11-16 39-66 18-30 A1 5 11-11 32-32 70-70
17-17 11-16 39-66 18-30 A1 6 11-11 32-32
70-70 17-17 11-16 39-66 18-30 A1 7 11-11
32-32 70-70 17-17 11-16 39-68 18-30
A1 8 11-11 32-32 70-70 17-17 11-16 39-68
18-29 A1 9 11-11 33-33 69-69 17-17 11-16
39-68 18-29 A1 10 11-11 33-33 69-69 17-17
11-16 39-68 18-29 A1 11 11-11 33-33 69-69
17-17 11-16 39-68 18-29 A1 12 11-11 33-33
71-71 18-18 11-16 39-68 18-29 A1 13 11-11
33-33 71-71 18-18 11-16 39-68
18-31 A1 14 11-11 33-33 71-71 18-18 11-16
39-68 18-31 A1 15 11-11 33-33 71-71 18-18
11-16 39-68 18-31 A1 16 11-11 32-32 71-71
18-18 11-16 39-68 18-29 A1 17 11-11
32-32 71-71 17-17 11-16 39-68 18-29
A1 18 11-11 32-32 69-69 17-17 11-16 39-68
18-29 A1 19 11-11 32-32 72-72 17-17 11-16
40-null 18-29 A2 20 11-11 9 32-32 72-72
19-19 11-16 40-null 18-29 A2
132
133
  • China Camp State Park (Marin) n24
  • Bean Creek Forest (Santa Cruz) n24
  • Nurseries n14

133
Write a Comment
User Comments (0)
About PowerShow.com