Cone and Seed Weight Relationships in Douglas-Fir From Western and Central Oregon Author(s) Frank C Sorensen and Richard S Miles Source Ecology Vol 59 No 4 (Summer 1978) pp 641-644 Published by Ecological Society of America Stable URL httpwwwjstororgstable1938763

Accessed 20012015 1947

Your use of the JSTOR archive indicates your acceptance of the Terms amp Conditions of Use available at httpwwwjstororgpageinfoaboutpoliciestermsjsp

JSTOR is a not-for-profit service that helps scholars researchers and students discover use and build upon a wide range of content in a trusted digital archive We use information technology and tools to increase productivity and facilitate new forms of scholarship For more information about JSTOR please contact supportjstororg

Ecological Society of America is collaborating with JSTOR to digitize preserve and extend access to Ecology

httpwwwjstororg

This content downloaded from 166610557 on Tue 20 Jan 2015 194727 PM All use subject to JSTOR Terms and Conditions

~cuugmiddot iiJ(4J Jl)7X pp 641-M4 c IIJ7X hy lht Fc1logical Society of America

CONE AND SEED WEIGHT RELATIONSHIPS IN DOUGLAS-FIR FROM WESTERN AND CENTRAL OREGON 1

fRANK C SORENSEN AND RICHARDS MILES USDA Forest Senbullice Pacific Northwest Forest and Range Experiment Station Forestry Sciences Lahoratorv 3200 Jejjrson Wav Corvallis Oregon 9733 USA

Ahstroct Cone sizes and weights and seed weights were determined for X9 trees in 9 stands in the Oregon Coast Ranges Cascade Ranges and central Oregon From west to east this represented a transition from mesic to xeric sites and from longer to shorter growing seasons Differences among regtons for all cone traits and regression coefficients for change in cone and seed wetghts with distance from the ocean were s ignificant Cone size and weight decreased seed weight increased and the regression coefficient of seed weight on cone weight based on trees within regions increased with distance from the ocean Geographic variation in cone and seed weights was compared with variation in the phenology of cone and seed development as reported in the literature This comparison indicated that where relatively la rge seeds were important to regeneration success and the growing season was short Douglas-fir shltJwed at least 3 changes in the timing of development as compared with milder sites (I) The cone growth period was reduced more than the seed growth period (2) The overlap in time of the developmental cycles of cone and seed was imreased (3) Cones a nd seeds grew during an increased portion of the growing season

Knmiddot 11 middotords Adaptation Dou~las-fir length o(grmling season naturaselection Ore~rJ11 mcial ariation trees

I NTRODUCTION

Seed size often increases with increasing dryness (Bake r 1972) presumably because of the need on dry sites for vigorous earl y seedling development In ma ny coniferous species seed size is also found to increase with increasing cone size (Willis a nd Hofmann 1915 Eliason and Heit 1940 Plym Forshell 1953 Squillace 1957 Florescu et al 1960 Dashkevitch 1961 Veracion 1964 Goggans a nd Posey 196H) Cone size in turn often increases with geographic changes that a re asshysociated with a n increase in the length of the growing season for e xample decreas ing latitude (Langle t 1938 Sziklai 1969) and decreasing e levation (Atay 1959 Hermann 1968)

In Oregon on the average elevation increases moi sture decreases and length of the growing season decreases from the Pacific Coast to the central part of the state This means that from west to east seed size would be expec ted to increase because of increasing dryness but to dec rease because of decreasing growshying season length and Jecreasing cone size We report here an analysis of seed sizecone size relationships in Douglas-fir ltPseudotsu~a menziesii) along a transect from the Coast Ranges to central Oregon

M ETHODS

Ten undamaged cones were taken at random from upper-c rown cone collections from ten 50- to 100-yearshyolu trees at 3 locations within each of the following regions (I) Coast Ra nges (2) CascaJe Ranges (upper elevation and east slopes) and (3) central Orego n (Table 1) Cone length anJ maximum width (measured with calipe rs to the nearest millimetre) dry weight

Manusnipt rclteivcd 7 June 1977 accepted 15 Decemshyber 1977

(after 72 hat 80degC) and number of scales (in the part of the cone bearing round seeds) were ueterntineu for each cone Seeds were ex tracted from the entire colshylection from each tree separateJ on an a ir column into filled and empty seeds and the number of filled seeds determined Seed weight was calculated from 100-seeu samples of filled seeds

One-hunureu-seeJ weights numbers of filled seeds pe r cone and weight of filled secus pe r gram of cone (dry weight) were ana lyzed using a nested classificashytion Sources of variation were regions seed sources within regions and trees within seed sources Cone traits were anal yzed similarly but inc luded cones within trees as an additional leve l of sampling

One-hunureu-seeu weight cone weight and filleushyseeu weight per gram of cone weight we re regresseJ on distance from the ocean and I 00-seeu weight and seed weight per gra m cone weight were regressed on cone dry weight Because variation among locations within regions was not significant the regress ions were based on regional means or on tree means within regions Within-region regress ion lines were compared using analysis of covariance afte r testing for homoshygeneity of variance (Sneuecor and Cochran 1967)

R ESULTS

Diffe re nces bet ween regions a nd bet ween trees within regions were highly significant for a ll cone tra its (Table 2 top) Differences between locations within regions were nonsignificant or in the case of cone scale number barely so a nd contributed o nl y a minor component to the variation The uniformit y within reshygions vis-a-vis between regions suggests a t first glance nonclinal differences between regions Howshyever because the sample poi nts within regions were close together and the regions large the result s also

This content downloaded from 166610557 on Tue 20 Jan 2015 194727 PM All use subject to JSTOR Terms and Conditions

642 FRANK C SORENSEN AND RICHARDS MILES Ecology Vol 59 No 4

TBII I Topographic data for the seed sources and climatic and growth data for the 3 regions Climatic data are averaged from representative weather stations in each region (Johnsgard 1963) Cone development periods which refer to time from pollen shed to seed shed are estimated from unpublished records of authors and from anonymous ( 1952) Ching and Ching ( 1962) and Wheeler ( 1967)

Potential evap-Cone

Distance Precipitation (mm) oration (mm)

development Elevation from ocean May- May- period

Localities (m) (km) Annual September September (wk)

Coast Ranges Tidewater 215 19 Oleman Creek 215 41 Marys Peak 365 50

i (region) 265 37 ]()50 250 190 20-2 1

Cascade Range Tombstone Pass 1525 154 Santiam Pass 1400 179 Metolius River 915 195

r (region) 1280 176 1200 180 225 17- 18

Central Oregon Grizzly Mountain 1555 250 Wildcat Mountain 1430 283 Baldy Mountain 1735 421

(region) 1573 318

would be consistent with a clinal pattern of variation Neither regions nor locat ions within regions differed significantly for seed weight traits

Cone size by all criteria decreased as distance from the ocean increased while 100-seed weight increased and weight of filled seeds per gram cone dry weight fluctuated irregularly Regression coefficients testing the response of cone weight and seed weight to disshytance from the ocean were both significant (p lt 05) and estimated with each 100-km increase in distance from the ocean at latitude 45degN an increase in I 00-seed weight of 05g and a decrease in cone weight of 97g

The regression coefficient of seed weight on cone weight using region means was not significant Regresshysion coefficients based o n trees within regions were positive being steepest for central Oregon and flattest for the Coast Ranges The coefficients differed signifshyicant ly among themselves (p lt 05) (Fig 1)

Cones from the 3 regions did not differ in number

140 125 300 12-14

of sound seeds per cone nor in weight of filled seeds produced per gram dry weight of cone When the latter trait was regressed against cone weight using region means and values for trees within region the calcushylated h-values were not significant but in all cases the weight of seeds per gram of cone weight decreased with increasing cone weight

DISCUSSION

Changes in seed and cone weights followed the exshypected trends Seed weight increased from west to east with increasing dryness of the si te and cone weight inc reased with increasing moisture and increasing lengt h of growing season Within a region (ie within areas of similar moisture conditions and growing seashyson length) the expected positive relationship between seed weight and cone weight was observed However between regions the seed we ightcone weight relationshyship was not positive When regeneration and growing

TAHIL 2 Components of variame expressed as percentages of total variation and significance for cone and seed traits Signifkance levels N S = nonsignificant middot = s ignificant at 5 level = significant at I level Seed weights were only available for individual trees not for cones in trees C V = coefficient of varia tion

Cones Seed wt Filled seed Scales ( 100 filled Fi lled seeds wtg cone

Component Length Width Wt (number) seed) cone WI

Regions 2l 9lt7o 304 270 45 6 96 NS 89fr NS ~7 NS

Sources in regions 00 NS 23 NS 36 NS 36 6 1 NS 8 1 NS 7 7 NS

Trees in sources 308 385 394 241 8437lt 83 (JC6 NS 896 Cones in trees 403 288 300 268 c v (r ) 117 7 9 2 12 II I 141 53 5 55 5

Mean values for regions Coast Ranges 692 em 353 em 802 g 47 2 1 16 g 175 262 mg Cascade Ranges 647cm 35 1 em 642 g 419 1 22 g 1 9 ~ 36 1 mg Centra l Oregon 547cm 300 em 530 g 362 130 g 122 295 mg

This content downloaded from 166610557 on Tue 20 Jan 2015 194727 PM All use subject to JSTOR Terms and Conditions

Summer 117H CONE AND SEED WEIGHTS IN DOUGLAS-FIR 643

15

14

E l5 u I ltJ w

sectu ()

0 g

11

CONE DRY WEIGHT (Grams)

FIG I Regressions of 100-seed weight on cone dry weight based on region means (solid line solid squares) and on individual trees within the 3 regions separately (dashed lines)

conJitions rcquireJ it the species was able to proJuce relativel y large seeJs in relatively small cones anJ in a relativel y short growing season Information in the literature on reproJuctive growth anJ phenology of Douglas-fir inJicates that this has been accomplisheJ primarily by aJjusting the timing of seeJ anJ cone Jeshyvelopment within the growing season anJ probably to a lesser extent by Jifferences in rates anJ Jurations of seeJ growth

Growth rates of Douglas-fir embryos have been reshycorJeJ at IXO anJ 915 min the Oregon Coast Ranges (Wheeler 1967) Temperature-Jevelopment rate relashytionships were remarkably similar at the 2 elevations as were the average temperatures Juring embryo growth (Wheeler 17 p 61) Average temperatures (Johnsga rJ 1963) for representative stations in the 3 regions we sampleJ Juring the expecteJ perioJs of embryo growth (June anJ Jul y in the Coast Ranges July anJ August in the CascaJ es and central Oregon) were 159 161 anJ 162degC respectively If the temshyperature-growth relationship observed by Wheeler ( 17) is ge neral then the occurrence of seeJs of simshyilar size would imply that the seed growth perioJs were also of similar length The fact that seed weight JiJ increase by = 10 from west to eas t inJicates that seeJ growth rates or growing season lengths may have varied somewhat between localities Nevertheless the observations suggest consiJerable uniformit y in rate anJ Juration of seeJ developme nt

However cone Jevelopment perioJs anJ vegetative phenology have been observeJ to Jiffer greatly both in length (Table I) and with respect to each other At low elevations in western Oregon Douglas-fir pollen flight is about miJ-April cones reach maximum Jry

weight in early August and seed shed starts in early September (Ching and Ching 1962) at upper elevashytions in the Cascades pollen shed does not occur until mid-May to early June anJ seeJ fall starts in mid-Sepshytember to early October (personal ohsenbullafions) At low elevations vegetative buJ flush occurs 6-7 weeks after floral bud flush (Sorensen and Campbell 1971) at high elevation the separation is 3-4 weeks (personal

ohsenbullafionl) At low elevation cambial growth is still strong at the time of seeJ sheJ at high elevation it is over by the time of seeJ fall (Emmingham 1977)

Taken together these observations indicate that where relatively large seeds a re advantageous and the growing season relatively short Douglas-fir has been able to adapt primarily because the length and the timshying of vegetative cone and seed growth perioJs could vary inJependently We propose specifically that unshyJer the conditions of increasing dryness and shortness of growing season (I) the cone growth period has deshycreased more than the seeJ growth period (2) the overlap in time of growth of cone and seeJ has inshycreased anJ (3) the proportion of the growing season used for cone anJ seed growth has increased

As a result of the first response variation in the length of growing season even of the cone growth period need not be accompanieJ by comparable varishyation in inJiviJual seed weight As a result of the secshyond anJ thirJ responses seeJs and cones Jevelop in common environmental conJitions for a larger portion of the growing season on sites with short seasons than on sites with long seasons Consequently seed size should be more closely relateJ to the cone size whemiddote the growing season is short

The thirJ response also suggests that if the growing season is short the time available for late summer seeJ and cone maturation may be much shortened For example collecting Douglas-Fir cones 2 weeks before natural cone opening may give cones anJ seeJs which are much more immature on sites with short seasons than on sites with long seasons

ACKNOWLEDGMEN TS

We are grateful for the he lpful comments and ideas of reshyv iewers a nd referees particularly R D Burdon and F T Ledig

LITERATURE C n ELJ

Anonymous 1952 Average day of year o n which given pheshynological development occurred in g iven species at points in Oregon a nd Washington Officia l Report Pacitic Northshywest Forest and Range Experime nt Station Portland Oregon USA

Atay I 1959 Research o n the seeds nf black pine (Pinus nig ru var pallugtiww) lnstanbul Univcrsit esi Orman Fakshyultesi Derg 9A4X- 96 (Turkish with Englis h su mmary )

Baker H G 1972 Seed weight in re latinn to e nvironmental conditions in California Ecology 53917-1010

C hing T M and K K Ching 1962 Physical and physshyiological c hanges in maturing Douglas-fir cones and seed Forest Science 8 2 1- 3 1

Dashkevich M D 196 1 Influence of cone size in common

This content downloaded from 166610557 on Tue 20 Jan 2015 194727 PM All use subject to JSTOR Terms and Conditions

644 FRANK C SORENSEN AND RI CHARDS MILES Ecology Vol 59 No4

pine on seed germination and quality and on growth of seedlings in progenies Lesnoj Zhurnal (Archangelsk) 4159-161 (In Russian)

Eliason E J and C E He it 1940 The s ize of Scotch pine cones as related to seed size and yie ld Journal of Forestry 3865-66

Emmingham W H 1977 Comparison of selected Douglasshyfir seed sources for cambial and leader growth patterns in four western Oregon environments Canadian Journal of Forest Research 7 154-164

Florescu I I Balan and S Carabela 1960 The fruiting of Larix decidua Mill in the Bucegi massif Revue Padshyurilor 75 264-268 (Forestry Abstract 22414)

Goggans J F and E C Posey 1968 Variation in seeds and ovulate cones of some species and varieties of Cashypressul Alabama Agricultural Experiment Station Circushylar No 160

Hermann R K 1968 Cone and seed variation of Douglasshyfir with elevation and aspect Proceedings of the Annual Meeting of the Western Forest Genetics Association Corshyvallis Oregon USA (Abstract only )

Johnsgard B A 1963 Tempera ture and water balance fo r Oregon weather stations Agricultural Experiment Stashytion Oregon State University Corvallis Oregon USA Special Report 150

Langlet 0 1938 Proveniensfcirsok med olika tradslag Svenska Skogsvardsfcireningens Tids krift 3655-278

Plym Forshell C 1953 The development of cones and seeds in the case of self- and cross-pollination in Pinus

svlvestris L Meddelanden fran Staten Skogsforskningsinshystitut 43nr 10 (Swedish with English summary)

Snedecor G W and W G Cochran 1967 Statistical methods 6th editio n Iowa State University Press Ames Iowa USA

Sorensen F C a nd R K Campbell 1971 Corre lation between dates of floral and vegetative bud flush in Douglasshyfir USDA Forest Service Pacific Northwest Forest and Range Experiment Station Portland Oregon Research Note PNW-143

Squillace A E 1957 Variations in cone properties seed yield and seed weight in western white pine when pollishynation is controlled School of Forestry Montana State University Missoula Montana USA Bulletin No5

Szikla i 0 1969 Preliminary notes on varia tion in cone and seed morphology of Douglas-fir P1eudotsuga menziesii (Mirb) Franco 2nd World Consultation on Forest Tree Breeding FAO- FO-FTB-69-69

Veracion V P 1964 Correlation of cone size and weight with the numbers size and weight of seeds of Benguet pine (Pinus insularis End) Philippines Bureau of Forestry Occasional Paper No 16

Wheeler W P 1967 Patterns of mitotic act ivity during late embryogeny germina tion and early seedling ontogeny of Pseudotsuga menziesii (Mirb) Franco Unpublished docshytoral thesis Syracuse University Syracuse New York USA

Willis C P and J V Hofmann 1915 A stud y of Douglasshyfir seed Proceedings of the Society of American Foresters 10141-164

This content downloaded from 166610557 on Tue 20 Jan 2015 194727 PM All use subject to JSTOR Terms and Conditions