Variación genética altitudinal entre poblaciones de Pinus pseudostrobus de Michoacán, México. Resultados de ensayo en casas de sombra en dos localidades.

Full text

Artículo Científico Rev. Fitotec. Mex. Vol. 35 (2): 111 –120, 2012
Recibido: 20 de Octubre del 2010.
Aceptado: 18 de Abril del 2012.
ALTITUDINAL GENETIC VARIATION AMONG Pinus pseudostrobus POPULATIONS FROM
MICHOACÁN, MÉXICO. TWO LOCATION SHADEHOUSE TEST RESULTS
VARIACIÓN GENÉTICA ALTITUDINAL ENTRE POBLACIONES DE Pinus pseudostrobus DE
MICHOACÁN, MÉXICO. RESULTADOS DE ENSAYO EN CASAS DE SOMBRA EN DOS LOCALIDADES
Cuauhtémoc Sáenz-Romero1*, Gerald E. Rehfeldt2, José Carmen Soto-Correa1, Selene Aguilar-Aguilar3,5,
Verónica Zamarripa-Morales3 and Javier López-Upton4
1Instituto de Investigaciones Agropecuarias y Forestales, Universidad Michoacana de San Nicolás de Hidalgo (IIAF-UMSNH). Km 9.5 Carretera Morelia-
Zinapécuaro. 58880, Tarímbaro, Michoacán, México. 2Forestry Sciences Laboratory, Mountain Research Station, USDA Forest Service. 1221 S. Main, Moscow,
Idaho 83843, USA. 3Facultad de Biología, Universidad Michoacana de San Nicolás de Hidalgo. Av. Francisco J. Mújica s/n, Col. Felícitas del Río. Morelia,
Michoacán, México. 4Forestal, Colegio de Postgraduados Campus Montecillo. Km. 36.5 Carr. México-Texcoco. 56230 Montecillo, Texcoco, Estado de México ,
México. 5Present address: Instituto de Ciencias Agrícolas, Universidad Autónoma de Baja California (ICA-UABC). Carretera a Delta s/n. 21705, Ejido Nuevo
León, Baja California, México.
*Corresponding autor (csaenzromero@gmail.com)
SUMMARY
Pinus pseudostrobus Lindl. is the forest species most important
economically in the state of Michoacán, at central-west México. We
investigated genetic variation among P. pseudostrobus populations
along an altitudinal gradient in the native indian community of
Nuevo San Juan Parangaricutiro, Michoacán, México. Cones were
collected from eight populations at 100 m of altitudinal intervals
between 2200 m and 2900 m. Seedlings were grown for approximately
two years in two shadehouse environments at Morelia, Michoacán,
México and at Moscow, Idaho, USA. Total height was periodically
measured during the second growing season to estimate a series of
growth indexes for both locations, and dry weights were obtained only
for the Morelia test. For the Moscow test there were significant
differences (P < 0.05) among populations for cessation of growth,
duration of the growth period, and final height. For the Morelia tests
there were differences among populations for foliage, shoot and total
dry weight (P < 0.025), and significant differences for stem dry weight
at P = 0.055 level. There was a strong altitudinal cline for dry weight
variables, with seedlings originated from populations from the lowest
altitudes having more biomass (r2 = 0.80, P = 0.003). Populations
separated altitudinally by about 295 m are likely to be genetically
different. We suggest delineation of three altitudinal seed zones (Zone
I: 2100 m to 2400 m; Zone II: 2400 m to 2700 m; Zone III: 2700 m to
3000 m), that can be translated into climatic zones delimited
according to mean annual temperatures or by degree days > 5 °C. To
accommodate climate change we suggest implementing assisted
migration programs by transferring populations 300 m upwards to
maintain adaptiveness of populations for future climates.
Index words: Pinus pseudostrobus, altitudinal genetic variation,
altitudinal seed zoning, assisted migration, climatic change.
RESUMEN
Pinus pseudostrobus Lindl. es la especie forestal económicamente
más importante en el Estado de Michoacán, al centro-oeste de México.
Se investigó la variación genética entre poblaciones de P.
pseudostrobus a lo largo de un gradiente altitudinal en los bosques de
la comunidad indígena de Nuevo San Juan Parangaricutiro,
Michoacán, México. Se colectaron conos de ocho poblaciones a
intervalos altitudinales de 100 m, entre 2200 m y 2900 m. Las plantas
se crecieron por aproximadamente dos años en dos diferentes
ambientes de casas de sombra: en Morelia, Michoacán, México y en
Moscow, Idaho, USA. La altura total se midió periódicamente durante
la segunda estación de crecimiento para estimar una serie de índices
de crecimiento para ambas localidades, y en Morelia se estimaron
variables de peso seco. En el ensayo de Moscow hubo diferencias
significativas (P < 0.05) entre poblaciones para terminación, duración
del período de crecimiento y para altura final. En el ensayo de Morelia
hubo diferencias significativas entre poblaciones para peso seco de
follaje, parte aérea y total (P < 0.025), y para peso seco del tallo al nivel
de P = 0.055. Existió un fuerte patrón clinal altitudinal para las
variables de peso seco, en donde las plantas originadas de poblaciones
de la menor altitud tuvieron mayores valores de biomasa (r2 = 0.80, P
= 0.003). Es probable que las poblaciones separadas por 295 m de
diferencia altitudinal sean genéticamente diferentes. Aquí se propone
la delimitación de tres zonas altitudinales (Zona I: 2100 m a 2400 m;
Zona II: 2400 m a 2700 m; Zona III: 2700 m a 3000 m), que se pueden
convertir en zonas climáticas delimitadas por temperatura media
anual o grados día > 5 °C. Para dar cabida al cambio climático, se
sugiere implementar programas de migración asistida para transferir
poblaciones hacia arriba (300 m), como una medida de adaptación de
las poblaciones a los climas futuros.

ALTITUDINAL GENETIC VARIATION IN Pinus pseudostrobus Rev. Fitotec. Mex. Vol. 35 (2), 2012
112
Palabras clave: Pinus pseudostrobus, cambio climático, migración
asistida, variación genética altitudinal, zonificación altitudinal de
semillas.
INTRODUCTION
Pinus pseudostrobus Lindl. is a pine tree that is
distributed mostly in the pine and pine-oak forests of
México. It commonly grows on volcanic soils and in
temperate to warm-temperate climates with annual
precipitations between 800 and 1500 mm (Perry, 1991;
Farjon and Styles, 1997; López-Upton, 2002). P.
pseudostrobus is known for its fast growth rates on good
quality sites, straight trunk and high wood quality
(López-Upton, 2002), all of which make this species
among the best species candidate for tree breeding and
extensive commercial plantations in México.
The native indigenous community from Nuevo San
Juan Parangaricutiro (NSJP), at Michoacán, western
México, is of the Purepecha ethnic group that practices
sustainable forest management on approximately 11 000
ha of pine forest. Their well-managed lands, owned under
a community property status, contrast with the
neighboring lands that have been largely deforested in
this economically poor region known as Purépecha
plateau (Sánchez-Pego, 1995; Jaffee, 1997). Forests tend to
be dominated by P. pseudostrobus, which is distributed
between 2200 to 2900 masl. Managers have applied
harvest practices seed tree in mature stands followed by
reforestation mostly with P. pseudostrobus seedlings
produced in a local nursery. Also, abandoned agricultural
fields are converted to P. pseudostrobus commercial forest
plantations. Seedlings produced by the nursery were
originated from cones collected within the community
forest, but at present there are no recognized seed zones.
Seedlings produced by the NSJP community are also
extensively planted throughout the neighboring
Purépecha plateau as part of governmental reforestation
programs.
Studies along altitudinal gradients, mostly in the
Rocky Mountains, USA, have shown that conifer
populations tend to be differentiated genetically in
response to differential selection pressures along
altitudinal gradients. In general, populations originated
from colder environments at higher altitudes, show lower
growth potential, shorter periods of shoot elongation, and
higher tolerance to freezing than populations originated
from mild environments at low altitudes, which tend to
have higher growth potentials, longer periods of shoot
elongation and higher freezing damage (Campbell, 1979;
Rehfeldt, 1988 1989 1991).
Unfortunately, forest tree populations soon will be
decoupled of the climate for which they are adapted.
Vegetation models suggest that by the end of the current
century, suitable climates for the conifer forests in the
Trans-Mexican Volcanic Belt could be reduced by 92 %
due to the ongoing climatic change (Rehfeldt et al., 2012).
These changes should result from projected temperatures
increasse by 3.7 °C and precipitation decreased by 18.2 %
in the average, by the end of the century in México
(Sáenz-Romero et al., 2010). If the climate to which P.
psedostrobus populations are adapted shifts to higher
altitudes, it is likely that current NSJP forests are going to
exhibit a decline. Such decline or die-off of large masses of
forest, apparently to causes related to climatic change, is
underway in many parts of the world (Hogg et al., 2002;
Breshears et al., 2005; Peñuelas et al., 2007; Worrall et al.,
2008; Rehfeldt et al., 2009; Allen et al., 2010; Mátyás, 2010;
Mátyás et al., 2010; Rehfeldt and Jaquish, 2010).
A previous study of Pinus pseudostrobus provenances
from an altitudinal gradient in the NSJP region was tested
in the field at two altitudinally contrasting sites, but
produced inconclusive results. It showed a very weak
pattern of altitudinal genetic differentiation among
populations for seedling height, with populations from
low altitude growing slightly more than those from higher
altitudes. However, the relatively small size of the field
tests, frost damage at the lower altitudinal field test site,
and damage caused by gophers, undoubtedly decreased
the statistical power so that genetic effects could not be
detected (Viveros-Viveros et al., 2005).
The objectives of this research were: (a) To determine
if an altitudinal pattern of genetic differentiation among P.
pseudostrobus natural populations exists on NSJP forests;
(b) To delimitate seed zones and establish seed and
seedling movement guidelines if altitudinal patterns are
detected; and (c) To predict the climate change impacts
for the locations where present P. psedostrobus
populations are currently distributed.

SÁENZ, REHFELDT, SOTO, AGUILAR, ZAMARRIPA Y LÓPEZ Rev. Fitotec. Mex. Vol. 35 (2), 2012
113
In order to achieve objective (a), we devised two
studies aimed at maximizing the expression of genetic
differences among populations and estimating genetic by
environment interactions by conducting two-year
common-garden shadehouse tests at environmentally
disparate locations in Morelia, Michoacán, México and
Moscow, Idaho, USA. Objective (b) was aimed toward
understanding the matching between genotypes and
environment, and thereby to increase the survivorship and
growth rate of P. pseudostrobus planted by the
community. Objective (c) is directed toward finding
management alternatives capable of maintaining
adaptativeness in future environments.
MATERIALS AND METHODS
Wind-pollinated cones were collected from
approximately eleven randomly selected trees from each
of eight P. pseudostrobus natural populations distributed
along an altitudinal gradient in the NSJP native indian
community forest of Michoacán, at central-west México.
Sampled populations were separated by an altitudinal
interval of approximately 100 m, from 2200 m (19° 27.8’
N, 102° 08.9’ W) to 2910 m (19° 28.4’ N, 102° 11.0’ W.
Note that revisiting the field sites prompted a correction
in population altitude of +100 m from those reported in
Viveros-Viveros et al. (2005; 2006). Average geographic
distance between contiguous populations was
approximately 0.6 km. Seeds from individual trees were
mixed by population. The trees represented by these
samples are termed populations while the location of a
population is called the provenance.
For the Moscow test, seedlings were grown in Spencer-
Lamaire© 750 cm3 pots on a commercial substrate. Seeds
were germinated inside a greenhouse where the seedlings
remained until early summer in their first year. Then they
were moved to a shadehouse (50 % shade) for the
summer, and returned to the greenhouse for the winter;
seedlings were subsequently moved back to the
shadehouse in early March of the second year where they
remained until growth ceased in the autumn. The
experimental design was a randomized complete block
design, with three blocks, eight provenances, and nine
seedlings per plot. Seedlings were watered as needed. Total
seedling height (mm) was measured at two-week intervals
during the second year, starting on April 3rd before shoot
elongation started and ending on October 1st when shoot
elongation had ceased.
For the Morelia test, seedlings were grown in 380 cm3
Broadway Plastics de México® pots on commercial
substrate Creciroot® for one year. Then, seedlings were
transplanted to a rectangular wooden-structure raised bed
filled with a 40 cm layer of 4:1 mix of Creciroot® substrate
and a local Andosol forest soil which was placed over a 20
cm layer of a extrusive volcanic small stones for improving
drainage. The raised nursery bed was build inside a
shadehouse (50 % shade). The experimental design was a
randomized complete block, with three blocks, eight
provenances and six seedlings in row plots within each
block. Seedlings were spaced 13 cm within plots and 17
cm among plots. The first and the last plots were flanked
by a protection row from randomly chosen seedlings.
Seedlings were watered as needed, but little irrigation was
required during the June-October rainy season.
Total seedling height (mm) was measured weekly
during the second year from January 8th before shoot
elongation started, to October 2nd when growth had
ceased, and seedlings were approximately 2 years of age.
Measurements at both locations, therefore, were made
during the second of two growing seasons. Seedlings at the
Morelia test were harvested when they were two years-old;
needles, branches, stem and roots were separated, dried
for 62 h at 60 ºC, and then weighed. Dry weights were also
expressed as derived variables: shoot (shoot = needles +
branches + stem), shoot:root ratio, and total dry weight.
Dry weight of seedlings growing at Moscow was not
recorded.
A modified logistic growth function for total height
was fit for each individual seedling on separate analysis for
each location (Moscow and Morelia), using PROC NLIN
of SAS (1999):
Yi = 1 / (1 + e (
β
0 +
β
1 X + (
β
2 / X))) [Eqn. 1]
where Yi = observation on the ith seedling (total height);
β
0,
β
1 and
β
2 are regression parameters; and X =
measurement date (Julian day).
Regression parameters (
β
0,
β
1 and
β
2) were used to
estimate a growth curve of predicted values for each
individual seedling, using the following model:

ALTITUDINAL GENETIC VARIATION IN Pinus pseudostrobus Rev. Fitotec. Mex. Vol. 35 (2), 2012
114
Pi = (1 / (1 + e (
β
0 +
β
1 X +
β
2 (1 / X)) )) Z [Eqn. 2]
where Pi = predicted growth (total height) for the ith
seedling;
β
0,
β
1 and
β
2 are regression parameters; X =
measurement date (Julian day); and Z = total elongation
(mm).
Variables used in analyses of genetic variation include
total elongation, the difference between the final
measurement of two-year seedling height and the initial
seedling height obtained at the beginning of the second
year of growth. The regression models were used also for
estimating the day during the second growing season on
which 2 mm of growth had occurred, for each seedling,
that is, the start of growth period in Julian days; the day on
which all but 2 mm of growth had occurred, that is, the
end or cessation of growth period in Julian days; the rate
of elongation between 20 % and 80 % of total elongation,
that is, the maximum growth rate; and the number of days
between start of growth and end of growth, the duration
of growth.
These variables were used in an analysis of variance to
test significance among populations, using PROC GLM of
SAS (SAS Institute, 1999). Ratio of variance component to
total variance was estimated using PROC VARCOMP
METHOD = REML (SAS Institute, 1999). These analyses
used the following statistical model:
Yijkl =
µ
+ Li + Bj(Li) + Pk + Li*Pk + Pk*Bj(Li) +
ε
ijkl
[Eqn. 3]
where Yijkl = observation on the lth seedling of the kth
population of the jth block in the ith location,
µ
= overall
mean, Li = effect of the ith location, Bj(Li) = effect of jth
block nested in the ith location, Pk effect of the kth
population, Li x Pj = interaction of location by population,
PkxBj(Li) = interaction of population by block nested in
location, and
ε
ijkl = error term; i=1,...s, j=1,...b, and
k=1,...t, and l = 1,…n, where s = 2, b = 3, t = 8, n = 9 in
Moscow test and n = 6 in Morelia test, which are the
number of locations, blocks, populations, and seedlings-
per-plot, respectively.
In addition, separate analyses were conducted for
variables measured at each location, including the dry
weights of seedlings grown at Morelia, with the following
model:
Yijk = µ + Bi + Pj + Bi*Pj + εijk [Eqn. 4]
where Yijk = observation on the kth seedling of the jth
population of the ith block, µ = overall mean, Bi = effect of
ith block nested, Pj effect of the jth population, BixPj =
interaction of population by block, and
ε
ijk = error term.
The relationship between the altitude of the seed
source and genetic variation among populations for those
variables for which population effects were significant was
assessed with linear and quadratic models, using PROC
REG (SAS, 1999). The linear model was:
Yij = β0 + β1Xi + εij [Eqn. 5]
and the quadratic model was:
Yij = β0 + β1Xi + β2Xi2 + εij [Eqn. 6]
where Yij = population mean,
β
0 = intercept,
β
1 and
β
2 =
regression parameters, Xi = altitude (m) of ith population
origin,
ε
ij = error term.
Seed zoning
Differentiation along the cline was interpreted relative
to the least significant difference (LSD, α = 0.20) in
altitude that must separate populations before one can be
reasonably certain of genetic differentiation (Rehfeldt,
1991; Sáenz-Romero et al., 2006). Provisional altitudinal
seed zones were delimitated using LSD.
Estimation of climatic variables
Mean annual temperature, annual precipitation,
annual degree days (> 5 °C), an annual aridity index (ratio
of square root of annual degree days to annual
precipitation), and several additional climate variables
were estimated for a total of eight Pinus pseudostrobus
localities for contemporary climate (1961 – 1990). Climate
estimations were obtained from spline climate surfaces
fitted from monthly average temperatures (mean,
maximum and minimum) and monthly precipitations
from numerous weather stations (Sáenz-Romero et al.,
2010).
Forecasted climate change estimates for the
provenances for the decade centered in year 2030, were

SÁENZ, REHFELDT, SOTO, AGUILAR, ZAMARRIPA Y LÓPEZ Rev. Fitotec. Mex. Vol. 35 (2), 2012
115
obtained after refitting the spline climate surfaces with
outputs of one global circulation model (Canadian Center
for Climate Modeling and Analysis, CCCMA), and one
emission scenario A2 (see details in Sáenz-Romero et al.,
2010). Point estimates for each provenance were obtained
by interrogating the spline climatic surfaces
(contemporary and year 2030) by using a web-based
interface (Crookston, 2010).
Climatic clines in genetic variation among
populations were assessed with regression models using
PROC REG (SAS Institute, 1999) of genetic responses of
populations on provenance climate variables. The clines
were examined in the context of the seed and seedling
transfers required for realigning genotypes and climate
for the year 2030.
RESULTS AND DISCUSSION
Differences among test locations
Analyses of variance detected significant statistical
differences between locations (P ≤ 0.01) for the start, end,
duration and amount of shoot elongation. Final height
was significant at P = 0.0527 level, although there was no
significant difference between locations for growth rate
(Table 1). Also, there were no significant genotype by
environment interactions (Location x Provenance).
In comparison to provenances growing at Moscow,
those at Morelia showed a higher total elongation (412
mm vs. 308), an earlier start date (day 47 vs. day 113),
earlier ending date (day 227 vs. day 250), longer duration
of shoot growth (181 days vs. 138) and larger final height
(571 mm vs. 512) (Figure 1). Similar result were found for
6-month-old seedlings of Pinus patula provenances, also
originated from an altitudinal gradient and growing in
nursery conditions in contrasting localities (Sáenz-
Romero et al., 2011). The best growth occurred at the
locality providing the best environmental conditions, and
all populations responded similarly to the favorable
conditions so that there were no detectable interactions of
genotype by environment.
Differences among populations
When analyses used data from both sites
simultaneously, differences among populations were not
significant for any seedlings trait related to seedling
growth in height (start, end, duration, rate and amount of
elongation and final height), as shown in Table 1.
However, when the analyses were conducted separately
for each location, significant differences were detected
among populations in the Moscow test for the ending date
and duration of elongation and for final height.
Differences for total elongation also were significant at P =
0.076 (Table 1). In contrast, at Morelia, no significant
differences among populations were detected for all traits
related to shoot growth (Table 1). Similar results were
found for the same provenances of the same age growing
in field conditions at two localities: since some traits
showed significant differences among provenances in one
field location and not on the other, while differences in
other traits could not be detected at either location
(Viveros-Viveros et al., 2005).
Dry weight traits at the Morelia test, however, showed
large differences among populations for needle, shoot and
total dry weight (P < 0.025), while differences for stem dry
weight were significant at P = 0.055. No significant
differences were detected for dry weight of branches and
roots or for the shoot:root ratio dry weight (Table 2).
Altitudinal pattern of genetic differentiation
Population means for traits measured at Moscow and
shown by ANOVA to be, significant were poorly related to
provenance altitude. The best fitting regressions were
obtained with the quadratic model, but the regressions
were not statistically significant, for the three traits: end of
growth (R2 = 0.49, P = 0.182), duration of growth (R2 =
0.28, P = 0.436), and seedling height (R2 = 0.10, P = 0.761).
A factor contributing to this poor fit of the regression
model was that the population originated at 2600 m had
an atypical low mean value expected for its altitude
(Figure 2). Thus, in general, populations from middle
altitudes achieved a better growth through a longer
duration of shoot elongation (scatter plot not shown, but
mirrors that of Figure 2), while populations from the
extreme low and high altitudinal limits showed less
growth. The same trend was found for 6-month-old
seedlings from P. patula provenances: mid-altitude
provenances had better growth than provenances from
both extremes of altitudinal distribution, and the
quadratic regression model was not significant either
(Sáenz-Romero, et al., 2011).

ALTITUDINAL GENETIC VARIATION IN Pinus pseudostrobus Rev. Fitotec. Mex. Vol. 35 (2), 2012
116
In contrast, a clear altitudinal pattern was revealed for
dry weight traits, with populations from lower altitudes
having larger dry weights than populations from higher
altitudes (Figure 3). The fit of regression models of
population average dry weight on provenance altitude for
dry weight traits were highly significant for both linear
regression and quadratic models. The fit of the linear
regression models were, needle dry weight (R2 = 0.75, P =
0.005), stem dry weight (R2 = 0.69, P = 0.011), total dry
weight (R2 = 0.80, P = 0.003), and shoot dry weight (R2 =
0.72, P = 0.007). Figure 3 shows the behavior of total dry
weight; the scattering of data points was similar for needle,
stem and aerial dry weight. This pattern of altitudinal
genetic variation is similar to that found for P. oocarpa
(Sáenz-Romero et al., 2006) and for P. hartwegii (Viveros-
Viveros et al., 2009) in Mexican mountains.
Table 1. Two-location and by each location analysis of variance for two-year-old Pinus pseudostrobus provenance test.
Percent of contribution to total variance (%) and significance values (P).
S.V. d.f. Elongation Start End Duration Rate Height
% P % P % P % P % P % P
Two- location analysis
Location 1 32.5 .0027 96.0 .0001 27.9 .0069 53.7 .0009 23.1 .5320 9.8 .0527
Block(L) 4 1.3 .1933 0.0 .2669 0.0 .8082 0.0 .7344 0.0 .7637 0.0 .8384
Provenance 7 0.0 .7401 0.0 .5673 0.0 .8833 0.0 .8826 2.2 .0963 1.5 .3790
LxP 7 1.9 .1513 0.0 .4958 1.5 .0998 0.3 .1357 0.0 .7230 3.8 .2557
PxB(L) 27 1.2 .1785 0.0 .8549 0.0 .2761 0.0 .5133 4.5 .0333 1.6 .1251
Error † 63.2 4.0 70.6 46.0 70.2 83.3
Moscow, Idaho, USA
Block 2 0.0 .5125 0.2 .2139 0.0 .9105 0.0 .5138 0.0 .6282 0.0 .4302
Provenance 7 7.7 .0760 0.0 .5267 8.1 .0072 4.1 .0467 0.0 .5573 13.7 .0119
BloxProv 14 5.4 .0699 0.0 .6674 0.0 .7535 0.0 .7043 10.3 .0085 3.1 .1535
Error 185 86.9 99.8 91.9 95.9 89.7 83.2
Morelia, Michoacán, México
Block 2 2.5 .1725 0.0 .4020 0.0 .6656 0.0 .6510 0.0 .6567 0.0 .9164
Provenance 7 0.0 .5666 0.0 .5458 0.0 .6145 0.0 .6356 4.8 .2515 0.0 .7553
BloxProv 13 0.0 .5862 0.0 .8436 0.0 .5434 0.0 .6860 0.0 .4582 0.0 .4218
Error †† 97.5 100 100 100 95.2 95.2
Error degrees of freedom are: † 278 for elongation and start, and 275 for end, duration, rate and height; †† 93 for elongation and start and 90 for end, duration,
rate and height. S.V. = source of variation; d.f. = degrees of freedom.
Table 2. Analysis of variance for dry weight traits for two-year-old Pinus pseudostrobus provenances tested at Morelia,
Michoacán, México. Percent of contribution to total variance (%) and significance values (P).
S.V. d.f. Needles Branches Stem Root Total Aerial Aerial/Root
% P % P % P % P % P % P % P
Block 2 0.00 .3704 0.00 .7428 2.92 .0797 0.00 .5989 0.10 .1552 0.65 .1245 1.00 .5162
Provenance 7 3.96
.
0194
0.24 .2114 6.60
.
0545
0.32 .1561 4.87 .0225
6.61 .0141 0.00 .5390
BloxProv 13 0.00 .9621 0.00 .8180 0.00 .6881 0.00 .8444 0.00 .9255 0.00 .9166 7.57 .1257
rror 105 96.04 99.76 90.47 99.68 95.03 92.74 91.43
S.V. = source of variation; d.f. = degrees of freedom.

SÁENZ, REHFELDT, SOTO, AGUILAR, ZAMARRIPA Y LÓPEZ Rev. Fitotec. Mex. Vol. 35 (2), 2012
117
Figure 1. Average by location of elongation, date of growth start (Julian day),
date of growth end (Julian day), duration of shoot growth and final height (two-
year-old) for a Pinus pseudostrobus provenance test.
240
242
244
246
248
250
252
254
256
258
260
2100 2200 2300 2400 2500 2600 2700 2800 2900 3000
End of growth (days)
Altitude (m)
Figure 2. Population averages for end of growth from
Moscow test fitted against provenance altitude, using a
quadratic regression model (R2 = 0.49, P = 0.182).
60
65
70
75
80
85
90
2100 2200 2300 2400 2500 2600 2700 28 00 2900 3000
Total dr y weight (g)
Altitude (m)
Figure 3. Population averages for total dry weight from
Morelia test fitted against provenance altitude, using a
linear regression model (R2 = 0.80, P = 0.003).
512
138
250
113
308
571
181
227
47
412
0
100
200
300
400
500
600
Height 2 years
(mm)
Duration (days)
End (day)
Start (day)
Elongation (mm)
Trait
Plant growth (mm or days)
Morelia, Michoacan,
Moscow, Idaho,

ALTITUDINAL GENETIC VARIATION IN Pinus pseudostrobus Rev. Fitotec. Mex. Vol. 35 (2), 2012
118
Seed zoning and guidelines for seed movement
The least significant difference (LSD, α = 0.20)
between two populations was 10.93 g for total dry weight.
The ratio of LSD to the regression coefficient of
population means on altitude of the seed source suggests
that populations separated by 296 m are likely to be
genetically different. Considering that the maximum
altitudinal interval of the natural distribution of P.
pseudostrobus in the region of study is 700 m, that is, from
2200 to 2900 m, three seed zones would cover the entire
natural distribution. Thus, we used a 300 m of altitudinal
difference to design seed zones for P. pseudostrobus in the
NSJP, Michoacán region (Table 3), with delimitation
beginning at an altitude of 2100. Notice than an equivalent
criteria would be an altitudinal interval of ± 150 m from a
specific seed source.
Suitable guidelines for reforestation of ecological
restoration could be: (a) Reforestation of a given seed zone
using seedlings originating from the same seed zone, or
alternatively, (b) Reforest a site at a given altitude using
seedlings originating from seed collected from ± 150 m in
altitude from the site to be reforested.
For tree breeding and for establishing commercial
plantations in a particular seed zone, the guideline is to
use seeds of the best growing provenance or families
within the seed Zone with the provenances with best
performance. Among seed zones, Zone II (Table 3)
contains the populations with highest growth potential,
most particularly which originated at 2500 m (Figure 3).
The provenance from 2200 m had also a high mean dry
weight (Figure 3), but it had a low performance in seedling
height. The field performance of the 2500 m provenance
(Zone II) was also superior among three provenances
from NSJP and among several other provenances from
Michoacán. That provenance is from Cerro de
Tumbiscatillo, and it was incorrectly recorded in altitude
as 2400 m by Viveros-Viveros et al. (2005; 2006).
Climatic zoning
Seed zone limits can be translated to contemporary
climates by means of the strong association (indicated by a
regression analysis) between total dry weight and two
temperature variables that parallel our altitudinal
gradient: mean annual temperature (R2 = 0.81, P = 0.002)
and growing degree days (R2 = 0.81, P = 0.002). However,
association of total dry weight with precipitation is not
significant (R2 = 0.11, P = 0.427), largely because our
estimates of precipitation for this region show that it is
relatively constant across the altitudinal gradient. Annual
aridity index is significant (R2 = 0.68, P = 0.012),
apparently as result of the strong altitudinal cline of
growing degree days. Seed zone delineation based on
temperature variables followed the same procedure used
above for elevation. Additionally, we estimated the
association between elevation and mean annual
temperature (R2 = 0.99, P < 0.001) and growing degree
days (R2 = 0.99, P < 0.001) in order to relate altitudinal
seed zoning with the climatic seed zones (Table 3).
Our practical guidelines for designing seed zones
should be viewed as provisional. Future research needs to
be done on the correlation between dry weight at young
trees and traits of economic importance at later ages.
Nonetheless, the practical use of the current results for the
studied region is highly recommended, largely because
there are not guidelines for this species in México.
Climatic change predictions and assisted migration
strategies
Guidelines for adaping forest management to climatic
change, that is, goals aimed toward assuring that in 2030
genotypes will occupy climates similar to those they
inhabit today, must accommodate considerable
uncertainty. It is well known that temperatures decrease as
altitudes increase according to well established lapse rates.
Genetic responses parallel this trend (Figure 4, solid
markers). For the seed zones of Table 3, zones differ by
about 0.88 °C in mean annual temperature and about 320
degree-days > 5°C. Our estimates for 2030 show mean
annual temperature to increase by about 1.6 °C and for
degree-days to increase by about 565. This means that the
climates now inhabited by the P. pseudstrobus populations
tested in this study should occur in 2030 at approximately
535 m of higher altitude than they occur today. Because
the climate may continue to warm, we recommend today
the transfer of seeds about 300 m, that is, from a lower
seed zone to a higher.
This strategy for accommodating the changing climate
would be a temporary guideline for assisting the migration
of P. pseudostrobus while maintaining adaptation to

SÁENZ, REHFELDT, SOTO, AGUILAR, ZAMARRIPA Y LÓPEZ Rev. Fitotec. Mex. Vol. 35 (2), 2012
119
changing temperatures (Sáenz-Romero et al., 2010). More
information is needed on responses of P. pseudostrobus to
climate and on the impacts of climate to contemporary
forests before concrete guidelines can be formulated.
In regard to climate change, it is also noteworthy that
the climate is projected to decline by about 170 mm by
2030 across this region. A decline in precipitation coupled
with increasing temperatures means that the climate will
become more and more arid. Our calculations of the
annual aridity index, a ratio of degree-days > 5 °C to
annual precipitation, show that the index should increase
but yet remain well within the limits inhabited by P.
pseudostrobus today. Nonetheless, an increase in aridity
will undoubtedly mean that forests will be of lesser density
and anticipated growth rates may not be achieved.
A more rigorous analysis of global warming impacts is
needed to articulate management strategies. To find a
more detailed solution might require modeling the
suitable habitat for contemporary and for future climatic
conditions, as it was done using the Random Forest
Analysis for P. chiapensis (Sáenz-Romero et al., 2010).
However, such an approach would require a more
sophisticated analysis than ours. In the meantime, we
suggest to move middle and upper provenances upwards
300 m, the maximum width of a seed zone. To transfer
them beyond the maximum width of a seed zone in
anticipation of the warming expected after 2030, would
increase the risk of frost damage in contemporary times
(Viveros-Viveros et al., 2007; Sáenz-Romero and Tapia-
Olivares, 2008).
CONCLUSIONS
Statistical analyses detected significant genetic
differentiation among P. psedostrobus populations, with a
strong altitudinal cline by which seedlings originating
from populations at low altitudes had larger biomass
values than populations from higher altitudes.
Populations separated by about 295 m of altitudinal are
expected to differ genetically. Thus, we suggest three
altitudinal seed zones (Zone I: 2100 m to 2400 m; Zone II:
2400 m to 2700 m; Zone III: 2700 m to 3000 m) that can
be translated to climatic zones delimited by mean annual
temperatures (Zone I: 15.50 to 14.62 °C; II: 14.62 to
13.75 °C; III: 13.75 to 12.87 °C) or by degree days > 5 °C
(Zone I: 3822 to 3506; II: 3506 to 3189; III: 3189 to 2872).
As a general response to climatic change, we suggest to
implement assisted migration altitudinally upwards (300
m) as a measure to maintain population adaptation by
realigning genotypes to future climates.
Figure 4. Provenance mean annual temperature of
contemporary and year 2030 climate plotted against
provenance altitude. Arrow indicates altitudinal
upward movement needed to match a temperature for
which a population is adapted at present, that will occur
at higher elevation in year 2030.
Table 3. Simplified limits, ranges and intervals of three Pinus pseudostrobus seed zones, based on intervals of altitude,
mean annual temperature and degree days > 5°C of contemporary climate.
Seed zone Altitude (m) Mean annual temperature (°C) Degree days > 5 °C
Limits (m) Range Interval
(±)
Limits (°C) Range Interval
(±)
Limits Range Interval
(±)
Lower Upper Lower Upper Lower Upper
1 2100 2400 300 150 15.50 14.62 0.88 0.44 3822 3506 317 158.5
2 2400 2700 300 150 14.62 13.75 0.88 0.44 3506 3189 317 158.5
3 2700 3000 300 150 13.75 12.87 0.88 0.44 3189 2872 317 158.5
12
13
14
15
16
17
2100 2200 2300 2400 2500 2600 2700 2800 2900 3000
Mean annual temperature (0C)
Altitude (m)
Contemporary
Year 2030

ALTITUDINAL GENETIC VARIATION IN Pinus pseudostrobus Rev. Fitotec. Mex. Vol. 35 (2), 2012
120
ACKNOLEDGEMENTS
Funding was provided to CSR by Mexican Council of
Science and Technology and the Mexican National
Forestry Commission (CONACYT-SIMORELOS-2000-
0306021 and CONACYT-CONAFOR-2002-C01-4655),
the State of Michoacán (CONACYT-MICHOACÁN-
2009-127128), the Coordinación de la Investigación
Científica of the Universidad Michoacana de San Nicolás
de Hidalgo (5.1) and the USDA- Forest Service, Rocky
Mountain Research Station (01-JV-11222063-183).
Thanks to Manuel Echeverría, Rafael Echeverría, Luis
Toral, Felipe Aguilar and other persons from the Forestry
Office of the native indian community of Nuevo San Juan
Parangaricutiro, Michoacán, and to Ernesto Moreno,
Daniel Saldívar, Víctor Quiñonez and other personnel of
the Michoacán State Forest Commission for their help
with seed collection. We thank Patrick Wells at Moscow
and Soraya González and Guadalupe Hernández at
Morelia for their help for experiments maintenance, and
to Hans Nienstaedt for valuable comments.
REFERENCES
Allen C D, A K Macalady, H Chenchouni, D Bachelet, N McDowell, M
Vennetier, T Kitzberger, A Rigling, D D Breshears, E H Hogg,
P Gonzalez, R Fensham, Z Zhangm, J Castro, N Demidova, J
H Lim, G Allard, S W Running, A Semerci, N Cobb (2010) A
global overview of drought and heat-induced tree mortality
reveals emerging climate change risks for forests. For. Ecol.
Manage. 259:660-684.
Breshears D D, N S Cobb, P M Rich, K P Price, C D Allen, R G Balice,
W H Romme, J H Kastens, M L Floyd, J Belnap, J J Anderson,
O B Myers, C W Meyer (2005) Regional vegetation die-off in
response to global-change-type drought. Proc. Nat. Acad. Sci.
102: 15144-15148.
Campbell R K (1979) Genecology of Douglas-fir in a watershed in the
Oregon Cascades. Ecology 60:1036-1050.
Crookston N L (2010) Custom Data Requests. In: Current and Future
Climate Estimates, Plant-Climate Relationships, and
Potential Impacts of Global Warming Scenarios in
Western North America and Mexico. Available at:
http://forest.moscowfsl.wsu.edu/climate/. Moscow Forestry
Sciences Laboratory, Rocky Mountain Research Station, USDA
Forest Service. (February 2010).
Farjon A, B T Styles (1997) Pinus (Pinaceae). Flora Neotropica
Monograph 75. New York Botanical Garden, New York, USA.
291 p.
Hogg E H, J P Brandt, B Kochtubajda (2002) Growth and dieback of
aspen forests in northwestern Alberta, Canada, in relation to
climate and insects. Can. J. For. Res. 32:823-832.
Jaffee D (1997) Restoration where people matter; reversing forest
degradation in Michoacán, México. Restor. Manage. Notes
15:147-155.
López-Upton J (2002) Pinus pseudostrobus Lindl. In: Tropical Tree Seed
Manual. J A Vozzo (ed). USDA Forest Service. Pp: 636-638.
Mátyás C (2010) Forecasts needed for retreating forests. Nature
464:1271.
Mátyás C, I Berki, B Czúcz, B Gálos, N Móricz, E Rasztovits (2010)
Future of beech in Southern Europe from the perspective of
evolutionary ecology. Acta Silv. Lignaria Hungarica 6:91-110.
Peñuelas J, R Oyaga, M Boada, A S Jump (2007) Migration, invasion
and decline: changes in recruitment and forest structure in a
warming-linked shift of European beech forest in Catalonia (NE
Spain). Ecography 30: 830-838.
Perry J P (1991) The Pines of México and Central America. Timber
Press, Portland Oregon, USA, 231 p.
Rehfeldt G E (1988) Ecological genetics of Pinus contorta from the
Rocky Mountains (USA): a synthesis. Silvae Genet. 37:131-135.
Rehfeldt G E (1989) Ecological adaptations in Douglas-fir (Pseudotsuga
menziesii var. glauca): a synthesis. For. Ecol. Manage. 28:203-
215.
Rehfeldt G E (1991) A model of genetic variation for Pinus ponderosa in
the Inland Northwest (U.S.A.): applications in gene resource
management. Can. J. For. Res. 21: 1491-1500.
Rehfeldt G E, D E Ferguson, N L Crookston (2009) Aspen, climate, and
sudden decline in western USA. For. Ecol. Manage. 258:2353-
2364.
Rehfeldt G E, B C Jaquish (2010) Ecological impacts and management
strategies for western larch in the face of climate-change. Mitig.
Adapt. Strategies Global Change 15:283-306.
Rehfeldt G E, N L Crookston, C Sáenz-Romero, E Campbell (2012)
North American vegetation model for land-use planning in a
changing climate: a solution to large classification problems.
Ecol. Applica. 22:119-141.
Sáenz-Romero C, R Guzmán-Reyna, G E Rehfeldt (2006) Altitudinal
genetic variation among Pinus oocarpa populations in
Michoacán, México; implications for seed zoning, conservation
of forest genetic resources, tree breeding and global warming.
For. Ecol. Manage. 229:340-350.
Sáenz-Romero C, B L Tapia-Olivares (2008) Genetic variation in frost
damage and seed zone delineation within an altitudinal transect
of Pinus devoniana (P. michoacana) in Mexico. Silvae Genet.
57:165-170.
Sáenz-Romero C, G E Rehfeldt, N L Crookston, P Duval, R St-Amant,
J Beaulieu, B A Richardson (2010) Spline models of
contemporary, 2030, 2060 and 2090 climates for México and
their use in understanding climate-change impacts on the
vegetation. Climatic Change 102:595-623.
Sáenz-Romero C, L F Ruiz-Talonia, J Beaulieu, N M Sánchez-Vargas,
G E Rehfeldt, (2011) Genetic variation among Pinus patula
populations along an altitudinal gradient. Two environment
nursery tests. Rev. Fitotec. Mex. 34:19-25.
Sánchez-Pego M A (1995) The forestry enterprise of the indigenous
community of Nuevo San Juan Parangaricutiro, Michoacán,
México. In: Proc. Symposium Forestry in the Americas:
Community Based Management and Sustainability. Institute for
Environmental Studies and Land Tenure Center, University of

SÁENZ, REHFELDT, SOTO, AGUILAR, ZAMARRIPA Y LÓPEZ Rev. Fitotec. Mex. Vol. 35 (2), 2012
121
Wisconsin-Madison. 3-4 February 1995, Madison, Wisconsin,
USA. pp: 173-196.
SAS Institute Inc. (1999) SAS/STAT Computer Software. Release 8. SAS
Institute Inc., North Carolina, USA. 3809 p.
Viveros-Viveros H, C Sáenz-Romero, J López-Upton, J J Vargas-
Hernández (2005) Variación genética altitudinal en el
crecimiento de plantas de Pinus pseudostrobus Lindl. en campo.
Agrociencia 39:575-587.
Viveros-Viveros H, C Sáenz-Romero, J J Vargas-Hernández, J López-
Upton (2006) Variación entre procedencias de Pinus
pseudostrobus establecidas en dos sitios en Michoacán, México.
Rev. Fitotec. Mex. 29:121-126.
Viveros-Viveros H, C Sáenz-Romero, J López-Upton, J J Vargas-
Hernández (2007) Growth and frost damage variation among
Pinus pseudostrobus, P. montezumae and P. hartwegii tested in
Michoacán, México. For. Ecol. Manage. 253:81-88.
Viveros-Viveros H, C Sáenz-Romero, J J Vargas-Hernández, J López-
Upton, G Ramírez-Valverde, A Santacruz-Varela (2009)
Altitudinal genetic variation in Pinus hartwegii Lindl. I.: height
growth, shoot phenology, and frost damage in seedlings. For.
Ecol. Manage. 257:836-842.
Worrall JJ, L Egeland, T Eager, R A Mask, E W Johnson, P A Kemp, W
D Shepperd (2008) Rapid mortality of Populus tremuloides in
southwestern Colorado, USA. For. Ecol. Manage. 255: 686-696.

11