Professional Documents
Culture Documents
DOI 10.1007/s11258-010-9737-6
Received: 2 September 2009 / Accepted: 12 February 2010 / Published online: 27 February 2010
Springer Science+Business Media B.V. 2010
123
54
123
environment is constant and analyze a single population for few time intervals (Valverde et al. 2004;
Valverde and Zavala-Hurtado 2006; Mandujano et al.
2007b). Most ecosystems, however, can be heterogeneous in both time and space showing strong
variation in their environmental conditions among
years (Mandujano et al. 2001, 2007a; Esparza-Olgun
et al. 2005). Recognizing the importance of this
heterogeneity on population behavior has resulted in
an increase in the number of studies that contrast the
population structure and dynamics between habitats
and time intervals (e.g. van Groenendael and Slim
1988; Svensson et al. 1993; Navarro and Guitian
2003; Esparza-Olgun et al. 2005; Mandujano et al.
2007a).
Traditionally, the population dynamics of plant
species have used transition matrix models (van
Groenendael and Slim 1988; Silvertown et al. 1993;
Svensson et al. 1993; Menges 2000; Caswell 2001,
Ramula 2008). These models constitute a powerful
analytical tool since they allow the projection of the
potential fate of a population under different theoretical scenarios, as well as the evaluation of the
relative contributions of the demographic processes
occurring in different life cycle stages (de Kroon
et al. 1986; Silvertown et al. 1993). The use of
stochastic matrix models, which allow the inclusion
of the spatial and temporal environmental variability
into the analysis of the long-term population dynamics (Caswell 2001; Mandujano et al. 2001; Pico and
Riba 2002; Valverde et al. 2004; Boyce et al. 2006),
has recently demonstrated that the effect of stochasticity on the long-term fate of populations may
depend on an interaction between life history traits
and sensitivity to environmental variance. In this
context, stochastic demographic analyses may contribute to our understanding of the causes of rarity, at
least in the aspect concerned with limitation in
population numbers and rates of seedling recruitment.
Many cacti populations have been subjected to
intense perturbations, such as habitat fragmentation,
illegal collection and trade, and land use change
toward farming and cattle ranching. Consequently,
several species are now facing marked population
declines that need urgent attention (Anderson et al.
1994; Carrillo-Angeles et al. 2005; Mandujano et al.
2007b). This need for information is in striking
contrast to the few number of species that have been
studied from a demographic point of view (Godnez-
55
Biosphere Reserve, Mexico, a floristic region recognized as a center of endemism and diversity of cacti
and other species (Villasenor et al. 1990). The species
grows in xerophilous scrub and dry tropical forest,
and populations are fragmented into small local
populations that are relatively far from each other.
Only seven populations are known to actually exist
(Peters and Martorell 2000), and they are established
on a total surface of 1058 ha, with a density ranging
from 450 to 6300 ind/ha. This study was carried out
from 1999 to 2004 (5 years of demographic data) in
the Cuicatlan Valley (17480 N, 96580 W) where
mean annual rainfall is 553 mm, most of which falls
between June and September, and a mean annual is
temperature of 25.5C. In 1999 when we started this
study, three populations were reported to exist, one of
which (Distrito de Etla) when revisited had already
been destroyed by urban development. The two
remaining populations were the ones reported in
herbarium specimens at the time (Voucher specimens
in Mexico City, Mexico MEXU 474935, 518191,
790174, 202345, 387911). The two populations were
living under contrasting disturbance conditions (Flores-Martnez and Manzanero 2005). The first site
(disturbed site, hereafter S1, 17480 4100 N,
96570 5000 W) is in an area fragmented by housing,
road construction, and sand-mine operations. The
second site (conserved site, hereafter S2,
17540 1900 N, 96570 5600 W) is 18 km north of S1
and still has large portions of native vegetation cover.
The sites also differed in plant density (900 ind/ha in
S1 and 3850 ind/ha in S2). The species is very
patchily distributed in its distributional range; so
extrapolations of total number of individuals are
always overestimates. Both sites have Neobuxbaumia
tetetzo (F.A.C. Weber) Backeb. as the dominant
species with associates Escontria chiotilla (F.A.C.
Weber) Rose, Pachycereus weberi (J.M. Coulter)
Backeb., Myrtillocactus geometrizans (Mart.) Console, Hechtia sp., Turnera difusa Willd, Plumeria
rubra L., and Bursera morelensis Ramrez. Individuals of M. huitzilopochtli can be found on both sides
of gorges running through the area, with no relation
of the number of individuals or their density with the
cardinal orientation of the slope. Since the beginning
of our study, individuals found in the vicinity of our
permanent plot at S1 were killed by the construction
of a secondary road leading out of the town of
Cuicatlan. We found no differences in rainfall
123
56
123
The contribution of the vital rates toward population growth was assessed through the construction of
sensitivity and elasticity matrices, a prospective
approach that helped us determine how k would be
affected by changes in survivorship, growth, and
fecundity (de Kroon et al. 1986; Mandujano et al.
2001; Boyce et al. 2006).
Life table response experiments (LTREs)
Life table response experiments (LTREs, Horvitz
et al. 1997; Caswell 2001) were used to estimate the
contribution of year, plant size, and site to the
variation of k in one analysis, and of year, site, and
demographic process to the variation of k the
elasticities of k in a second analysis. These analyses
provided a means of exploring the contribution of any
given entry from the transition matrix to the observed
value of k (Horvitz et al. 1997; Tickin and Nantel
2004; Mandujano et al. 2007a). LTREs have n
treatments (in this case n = 10, two levels of site
and five levels of year), yielding population projection matrices A(m) and rates of increase k(m) for
m = 1,,n. The main effects of each factor and the
interactions among factors in two-way cross-classified treatments can be decomposed. A(m) = A(ij) was
the transition matrix resulting from the ith level of
time (years, with five levels, 20002001, 20012002,
20022003, 20032004, and 20042005) and the jth
level of site (S1 or S2), and k(ij) its eigenvalue. A(r)
was the reference population projection matrix (mean
matrix, overall S1 and S2 by year combinations) with
corresponding rate of increase k(r). The effect of a
treatment m on k was then decomposed into contributions from each of the matrix entries (Caswell
2001).
Numerical simulations
We used numerical simulations to find out how
changes in demographic parameters would impact
population growth rates. In these models the population growth rate depends on different conditions
(annual matrices that have a certain probability of
occurrence), in such a way that a temporal sequence
of annual projection matrices is generated by a
stochastic process (Caswell 2001). To estimate ks (the
57
123
58
Results
We found that both sites showed some degree of
human disturbance (DI), but DI differed between
sites. The value of DI for the most disturbed site (S1
site) was 52.05, and 14.71 for the S2 site mainly
related to land degradation and human activities. This
disturbance was also reflected in plant density: mean
density was 90 ind/ha in S1 and 385 ind/ha in S2.
Population dynamics
We found that reproductive individuals do not
produce fruits in successive years. In S2, most
individuals produced fruits in one of the studied
years, while others produced fruits during two or
three consecutive years. No individuals, however,
produced fruits in all five studied years. On the
contrary, reproductive individuals at S1 only produced fruits once over the study period. The
percentage of plants bearing fruits was highly variable between years in S2 (7.71, 9.83, 25.74, 9.7, and
17.81 each year from 1999 to 2004, respectively) and
in S1 plants showed a decreasing fruit production in
= 0.7733 0.1220
C1
0.125
C2
0.5288
0.0214
0.2226
0.043
C3
0.5042
0.2414
0.1338
C4
0.5646
0.1126
0.0478
C5
0.6464
0.0144
0.0013
0.0008
0.0001
0.0018
= 0.8117 0.1122
C1
0.150
C2
0.3848
b
0.4876
0.0322
C3
0.569
0.245
0.095
C4
0.5484
0.181
C5
0.0908
0.6964
0.0058
0.002
123
0.0034
0.0068
0.0082
59
Table 1 Transition matrices by site and year for Mammillaria huitzilopochtli populations
C1
C2
C3
k = 0.8096
0.1535
C4
C1
0.0004
0.0005
0.005
0.002
C2
0.125
0.281
C3
0.351
0.538
0.054
78
C3
0.591
0.318
0.010
217
C4
0.321
0.568
0.040
37
C4
0.429
0.375
104
C5
0.013
0.216
0.760
25
mi
0.875
0.368
0.128
0.162
0.200
S1
20002001
S1
19992000
S2
C5
19992000
ni
0.0079
57
C1
C2
C3
k = 0.7356
0.1534
C4
C5
ni
C1
0.004
0.005
0.009
C2
0.150
0.236
0.017
110
C5
0.375
0.733
mi
0.850
0.173
0.244
0.240
0.267
C5
C1
C2
C3
k = 0.8832
0.1032
C4
C1
0.002
0.005
0.006
C2
0.150
0.731
0.096
45
C1
C2
C3
k = 0.6973
0.1554
C4
C1
0.003
0.001
C2
0.125
0.563
0.156
C3
0.063
0.578
0.447
0.107
64
C3
0.154
0.756
0.235
135
C4
0.031
0.340
0.143
47
C4
0.037
0.561
0.243
132
C5
0.043
0.536
28
74
mi
0.875
0.375
0.234
0.170
0.214
S1
20012002
C5
S2
C5
ni
0.004
16
20002001
C5
0.038
0.595
0.850
0.115
0.111
0.167
0.162
C5
C1
C2
C3
C1
0.0006
0.002
C2
0.125
0.500
C1
C2
C3
k = 0.8366
0.1241
C4
C1
0.008
0.013
0.015
C2
0.150
0.563
0.029
32
C3
0.300
0.435
0.043
62
C3
0.344
0.652
C4
0.355
0.565
23
C4
0.174
0.062
138
0.691
0.082
C5
0.130
0.647
17
C5
0.007
97
0.144
0.653
mi
0.875
0.200
0.210
0.261
0.353
mi
0.850
0.094
49
0.138
0.103
0.265
S1
20022003
C5
C5
S2
ni
0.0026
20012002
C1
C2
C3
k = 0.7338
0.2328
C4
C1
0.003
C2
0.125
0.500
C3
0.200
0.441
34
C3
0.682
0.519
C4
0.324
0.600
35
C4
0.398
C5
0.059
0.143
0.733
15
C5
0.009
mi
0.875
0.300
0.176
0.257
0.267
mi
0.850
0.091
0.065
S1
20032004
C5
C1
C2
C3
k = 0.8742
0.2154
C4
C1
0.0006
0.002
0.0026
C2
0.125
0.800
0.059
C3
0.200
0.529
0.125
C4
0.176
0.750
C5
0.031
mi
0.875
0.235
0.094
0.389
0.037
C1
C2
C3
C1
0.001
0.004
0.002
C2
0.150
0.227
0.009
22
0.042
108
0.547
95
0.295
0.872
47
0.116
0.128
ni
0.003
ni
k = 0.8729
0.1014
C4
S2
10
0.013
26
mi
k = 0.6556
0.5367
C4
20
ni
20022003
ni
0.007
C1
C2
C3
k = 0.7809
0.1422
C4
C1
0.002
0.007
0.009
C2
0.150
0.167
0.027
17
C3
0.667
0.600
0.126
75
0.056
32
C4
0.187
0.568
0.129
95
0.556
18
C5
0.013
0.053
0.629
70
mi
0.850
0.167
0.173
0.253
0.243
S2
ni
20032004
C5
ni
0.018
S1 = population disturbed site; S2 = population conserved site; C1 = seedlings \0.1 cm; C2 = juvenile of 0.12 cm; C3 = young adults of 2.1
4 cm; C4 = mature adults of 4.16 cm; and C5 = old adults of [6 cm, mi = mortality. In the first row, seedlings per class are given. Values in the
main diagonal indicate the probability of stasis, values below the diagonal are transitions corresponding to growth, and values above the diagonal
indicate regression. Instantaneous rates of growth (k SE) are given next to each site 9 year combination. ni is the number of individuals in each size
class
123
60
classes could be found between noncontiguous categories. Mortalities based on the total number of
individuals of the mean matrix were higher in S1 that
in S2 (35.9 and 30.3%, respectively) being variable
between years and sites (mi, Table 1). In both
populations the highest levels of mortality were
predominantly in size class C5, followed by C2 in S1
and C4 in S2 (Table 1).
Growth rates showed declining populations except
in S1 for 20012002 (k = 0.6556 0.5367) and
20032004 (k = 0.8742 0.2154, Table 1). However, the confidence intervals make the value of k at
equilibrium uncertain. The mean matrix showed
values of k below unity for both populations
(Fig. 1) suggesting that over the long term, populations are prone to be in further peril. We found
significant differences between the mean observed
and projected stable size distributions in both populations (Fig. 2, G-test; P \ 0.01). Differences were
mainly found by an excess of observed individuals in
size class C2 and C3 and a more than expected
number of individuals in the last two size classes.
Size classes C1 for both populations and C4 for S1
did not differ from projected stable size distributions.
123
Elasticity analysis
The elasticity analysis on mean annual matrices
showed that the elasticity of stasis was similar for
both sites (S1 = 73.7% and S2 = 75.4%), growth
elasticities were 14.11% and 12.52% for S1 and S2,
respectively, and the elasticity of fecundity was
higher in S2 (13.09%) than in S1 (11.8%). On a
yearly basis, the highest elasticity values corresponded to stasis in some years and to growth in
others suggesting that the populations respond to
precipitation by an increase in the relative importance
of growth (Table 1). In addition, the elasticity of
retrogression was higher at S1 (13.09%) than at S2
(11.8%). In terms of stasis, the highest values were
associated to size category C5 followed by C4 in S1
and by C3 in S2.
Life table response experiments (LTREs)
The LTREs showed different results between populations. In terms of size classes, S2 showed consistent
positive contributions from three out of four size
classes toward k, while in S1 all but size class C4
contributed negatively to k (Fig. 3a). When considering the effect of year (Fig. 3b), the contribution of
C1 was negligible for all years probably due the fact
that entries were estimated; however, contributions
were either consistently negative for size classes C3
or larger (19992000 and 20012002), and the
contributions for the other years are highly variable
between years. When subdividing the contributions
by demographic process, fecundity was consistently
negative for S1 and positive (except during 2000
2001) for S2 (Fig. 4). Stasis contributed negatively in
3 out of 5 years in S1 and 1 out of 5 in S2 (Fig. 4).
For growth, contributions in S1 and S2 tended to have
the same pattern except during 20012002, which
showed positive contributions in S1 and negative in
S2, negative contributions were very high in one year
(20002001, Fig. 4).
Numerical simulations
Mean matrix population growth rate in S1 was below
unity regardless of increases in fecundity, growth or
stasis (Fig. 5). The same decline was found for S2,
except that increases in fecundity above 50% gave
values above unity (Fig. 5). The stochastic growth
0.015
0.01
Contributions
Fig. 3 Contribution by a
size class (C1C5) from
analyzing the life table
response experiments
(LTREs) by population
(S1 = white bars and
S2 = black bars), and b by
size class and year (1999
2000 black, 20002001
white, 20012002 gray,
20022003 hatched, and
20032004 dotted), in order
from left to right. Positive
values are associated to
increases in the values
of k, while negative values
are responsible for a
decrease in k
61
0.005
0
-0.005
-0.01
-0.015
C1
C2
C3
C4
C5
Size class
0.03
Contributions
0.02
0.01
0
-0.01
-0.02
-0.03
C1
C2
C3
C4
C5
Size class
rates suggested that both populations are decreasing (S1, ks = 0.7711 0.0004 95% CI, S2, ks =
0.8094 0.0005 95% CI). The quasi-extinction
probability after 6 years was above 0.5 and reached
1 after 9 years for S1, while the population at S2 had
a quasi-extinction probability above 0.5 after
12 years and reached 1 after 17 years.
Discussion
The projected stable size distribution was significantly different from the observed distribution for
both populations as found in other demographic
studies (Rosas and Mandujano 2002; Esparza-Olgun
et al. 2002; Pico and Riba 2002, Godnez-Alvarez
et al. 2003). Interestingly, the reported pattern is
similar: an overrepresentation of the largest size
123
62
2002-2003
2001-2002
2000-2001
1999-2000
-0.002
-0.001
0.001
0.002
0.003
2003-2004
Year
2002-2003
2001-2002
2000-2001
1999-2000
-0.2
-0.1
0.1
0.2
0.3
G
2003-2004
2002-2003
2001-2002
2000-2001
1999-2000
-0.2
-0.15
-0.1
-0.05
0.05
Contributions
123
63
The best estimates of k suggest declining populations but the wide confidence intervals make this
conclusion very uncertain, both populations are
declining with one population tending to increase in
some years. The stochastic growth rate suggests a
dramatic story in which both populations are likely to
go extinct in the near future under current conditions.
The values of k did not correlate with high or low
seed production even though seeds do not generate a
seed bank (Flores-Martnez et al. 2008). Apparently,
perennial species can buffer environmental variation
through the rate of population growth, changing the
relative sensitivities of different demographic processes through time (Mandujano et al. 2001, 2007a;
Esparza-Olgun et al. 2002, 2005; Pico and Riba
2002).
The LTREs showed that contributions of C1
toward k were small compared to other demographic
processes, and a constant negative contribution of
fecundity toward k in all years. However, the
percentage of plants producing seeds in S2 was
always above 6% while plants in S1 had a continuous
decrease in the number of fruits (\5% for all studied
years) a pattern that was followed by all reproductive
categories. It is possible that human disturbance may
be influencing some aspects of reproduction. Being
self-incompatible, with no clonal propagation,
M. huitzilopochtli depends entirely on sexual reproduction (fruit and seeds) and seedling establishment,
two attributes that were more affected at S1, which
had higher disturbance. In rare and threatened species
of cacti, population decreases have been found to be
related to reduced investment in reproduction (Esparza-Olgun et al. 2005; Mandujano et al. 2007b),
which in turn is negatively effected by disturbance
(Jimenez-Sierra et al. 2007; Mandujano et al. 2007a).
The high elasticity values of stasis observed in both
populations suggest that they rely on the survival of
remaining adults, and therefore, population numbers
would decline as these adults disappear. Instead, the
S2 population had the highest elasticity values
associated to growth or stasis depending on the year
indicating a paramount importance of both demographic processes on the rate or population growth.
Under relatively benign conditions, changes in
growth rates affect k more intensely than under
disturbance regimes. If in addition to disturbance
individuals are collected (at a rate of 10% or more for
123
64
the largest size classes) in the disturbed site, populations will decrease over time.
Our results suggest that human disturbance may be
affecting population dynamics. Previous studies
applying the same disturbance technique have shown
that disturbance affects population density of Mammillaria pectinifera (Martorell and Peters 2005). Both
populations of M. huitzilopochtli under current conditions may be lost over time, especially if disturbance is not stopped and a means of increasing
seedling survival is found. The whole species can
only be found in seven sites some of which have
\450 individuals (Peters and Martorell 2000) jeopardizing the success of any recovery plan. The
demographic parameters of this species suggest that
its ability for population growth is severely limited by
a lack of reproduction, which was shown to have
important effects in reproductive populations.
Additionally, the high habitat specificity and small
distribution range of the species are attributes that
further make this species vulnerable to extinction. Few
studies have addressed endangered species in rupicolous habitats (Pico and Riba 2002; Garca et al. 2002;
Garca 2003); so there is little information on their
population dynamics. Specifically for the Cactaceae,
Martorell and Patino (2006) have suggested that some
of the species of the Series Supertextae included in the
Mammillaria (like M. huitzilopochtli and M. crucigera) that are susceptible to chronic disturbance were
once more widely distributed (Martorell and Peters
2005) and are currently restricted to remnant populations (e.g., Pico and Riba 2002). Poor recruitment, the
continuously decreasing fruit and seed production and
high mortality over time could be driving the decline
of M. huitzilopochtli populations. The results of this
study suggest three things: (1) This species is currently
undergoing declining populations. Over longer time
periods, we would expect the decline or extinction of
those populations having small population sizes, no
fecundity and under disturbed habitats; (2) Habitat
conditions are highly restrictive for this species,
limiting seedling survival, establishment of individuals and increasing mortality of adults; and (3) Even
though disturbance cannot be considered to have an
effect on population dynamics it does seems to have
some effect on reproduction and stasis, making the
population rate of increase highly dependent on
survival of individuals rather than other demographic
processes. Several species of the Cactaceae are
123
References
lvarez R, Godnez-A
lvarez H, Guzman U, Davila P (2004)
A
Aspectos ecologicos de dos cactaceas mexicanas amenazadas: implicaciones para su conservacion. Bol Soc Bot
Mex 75:716
Anderson EF (2001) The cactus family. Timber Press, Portland, Oregon, USA
Anderson EF, Arias-Montes S, Taylor NP (1994) Threatened
cacti of Mexico. Royal Botanical Gardens, Kew, England
Arias S, Guzman U, Mandujano MC, Soto GM, Golubov J
(2005) Las especies mexicanas de cactaceas en riesgo de
extincion I: Una comparacion entre listados NOM-059ECOL-2001 (Mexico), la lista roja (UICN) y CITES. Cact
Suc Mex 50:100125
Bowers JE (2000) Does Ferocactus wislizeni (Cactaceae) have
a between-year seed bank? J Arid Environ 45:197205
Boyce MS, Haridas CV, Lee CT (2006) Demography in an
increasingly variable world. Trends Ecol Evol 21:141148
Bravo Hollis H (1978) Las Cactaceas de Mexico. I. UNAM,
Mexico
Carrillo-Angeles IG, Golubov J, Rojas-Arechiga M, Mandujano MC (2005) Distribucion y estatus de conservacion de
Ferocactus robustus (Pfeiff.) Britton et. Rose. Cact Suc
Mex 50:3655
Caswell H (2001) Matrix population models: construction,
analysis and interpretation, 2nd edn. Massachusetts, USA
Clark-Tapia R, Mandujano MC, Valverde T, Mendoza A,
Molina-Freaner F (2005) How important is clonal
recruitment for population maintenance in rare plant
species? The case of the narrow endemic cactus, Stenocereus eruca, in Baja California, Mexico. Biol Conserv
124:123132
65
Jimenez-Sierra C, Mandujano MC, Eguiarte LE (2007) Are
populations of the candy barrel cactus (Echinocactus
platyacanthus) in the desert of Tehuacan, Mexico at risk?
Population projection matrix and life table response
analysis. Biol Conserv 135:278292
Kephart SR, Paladino C (1997) Demographic change ad
microhabitat variability in a grassland endemic, Silene
douglasii var, oraria (Caryophyllaceae). Am J Bot
84:179189
Mandujano MC, Montana C, Mendez I, Golubov J (1998) The
relative contributions of sexual reproduction and clonal
propagation in Opuntia rastrera from two habitats in the
Chihuahuan Desert. J Ecol 86:911921
Mandujano MC, Montana C, Franco M, Golubov J, FloresMartnez A (2001) Integration of demographic annual
variability in a clonal desert cactus. Ecology 82:344
359
Mandujano MC, Flores-Martnez A, Golubov J, Ezcurra E
(2002) Spatial distribution of three globose cacti in relation to different nurse-plant canopies and bare areas.
Southwest Nat 47:162168
Mandujano MC, Golubov J, Huenneke LF (2007a) Effect of
reproductive modes and environmental heterogeneity in
the population dynamics of a geographically widespread
clonal desert cactus. Popul Ecol 49:141153
Mandujano MC, Verhulst JAM, Carrillo-Angeles IG, Golubov
J (2007b) Population dynamics of Ariocarpus scaphirostris Bodeker (Cactaceae): evaluating the status of a
threatened species. Int J Plant Sci 168:10351044
Martorell C, Patino P (2006) Globose cacti (Mammillaria)
living on cliffs avoid high temperatures in a hot dryland of
Southern Mexico. J Arid Environ 67:541552
Martorell C, Peters E (2005) On the measurement of chronic
disturbance and its effects on the threatened cactus
Mammillaria pectinifera. Biol Conserv 124:199207
Menges E (2000) Population viability analysis in plants:
challenges and opportunities. Trends Ecol Evol 15:5156
Morris WF, Doak DF (2002) Quantitative conservation biology: theory and practice of population viability analysis.
Sinauer Associates, Sunderland, MA, USA
Mungua-Rosas MA, Sosa VJ (2008) Nurse plants versus nurse
objects: effects of woody plants and rocky cavities on the
recruitment of the Pilosocereus leucocephalus columnar
cactus. Ann Bot 101:175185
Navarro L, Guitian J (2003) Seed germination and seedling
survival of two threatened endemic species of the northwest Iberian Peninsula. Biol Conserv 109:313320
Peters E, Martorell C (2000) Conocimiento y conservacion de
las Mamillarias endemicas del Valle de Tehuacan-Cuicatlan. R166. CONABIO. Instituto de Ecologa, UNAM,
Mexico
Pico X, Riba M (2002) Regional-scale demography of Ramonda myconi: remnant population dynamics in a preglacial relict species. Plant Ecol 161:113
Rabinowitz D, Cairns S, Dillon T (1986) Seven kinds of rarity.
In: Soule ME (ed) Conservation biology. Sinauer, Sunderland, Mass, pp 182204
Ramula S (2008) Population dynamics of a monocarpic thistle:
simulated effects of reproductive timing and grazing of
flowering plants. Acta Oecol 33:231239
123
66
Rosas MD, Mandujano MC (2002) La diversidad de historias
de vida de cactaceas, aproximacion por el triangulo demografico. Cact Suc Mex 47:3341
Schemske DW, Husband BC, Ruckelhaus MH, Goodwillie C,
Parker IM, Bishop JG (1994) Evaluating approaches to
the conservation of rare and endangered plants. Ecology
75:584606
Schmalzel JR, Reichenbacher FW, Rutman S (1995) Demographic study of the rare Coryphantha robbinsorum
(Cactaceae) in Southeastern Arizona. Madrono 42:332348
SEMARNAT (2002) Norma Oficial Mexicana NOM-059ECOL-2001. Diario Oficial de la Federacion. 2a. seccion.
06 marzo 2002
Shreve F (1931) Physical conditions in sun and shade. Ecology
12:96104
Silvertown JW, Franco M, Pisanty I, Mendoza A (1993)
Comparative plant demographyrelative importance of
life cycle components to the finite rate of increase in
woody and herbaceous perennials. J Ecol 81:465476
Stubben C, Miiligan B (2007) Estimating and analyzing
demographic models using the popbio package in R. J Stat
Softw 22:123
Svensson BM, Carlsson BA, Karlsson PS, Nordell KO (1993)
Comparative long-term demography of three species of
Pinguicula. J Ecol 81:635645
123