Regeneration of California Oak Woodlands
2001-20051
Tara Barrett2 and Karen Waddell3
Abstract
The first (1981-1984) and second (1991-1994) statewide inventories of California’s oak
woodlands found low levels of regeneration for several common oak species. In 2001, a new
statewide inventory of California’s oak woodlands was initiated, with 10 percent of field plots
measured each year. The first five years of data (2001-2005) were used to examine
regeneration and sapling distributions in common oak woodland types. Blue oak (Quercus
douglasii Hook. & Arn.), valley oak (Quercus lobata Nee), and coast live oak (Quercus
agrifolia Nee) forest types are characterized by low numbers of saplings, having mean levels
of 60.5 (+/- 10.2), 22.2 (+/-9.8), and 77.8 (+/- 18.7) trees per acre for all species of trees in the
1 to 5 inches diameter breast height (DBH) class. In this third decade of monitoring, diameter
distributions for blue oak and valley oak departed from the expected inverse-J shape. Better
information on population-wide growth and mortality of saplings would be useful for
understanding whether these observations indicate trouble ahead for the long-term
sustainability of these oak species.
Keywords: California oak, forest monitoring, oak savannah, oak woodland, recruitment.
Introduction
The sustainability of oak woodlands is a matter of concern in California, due to issues
such as conversion of forest to developed land (Bolsinger 1988), the introduction of
Phytophthora ramorum and its effect on California black oak and coast live oak
(Rizzo and others 2002), and low levels of regeneration and sapling recruitment for
the more xeric oak forest types (Tyler and others 2002).
The low level of observable recruitment for some oak species has been
commented on over the decades (Sudworth 1908), although there is considerable
debate about whether it is truly a problem for long-term viability of oak woodlands
(Tyler and others 2006). Explanations for low levels of recruitment include
consumption by wildlife, livestock, and insects; environmental and chemical
inhibition from introduced grasses, and other environmental factors (McDonald
1990). While blue oak has been the focus of the bulk of research on regeneration,
other species including valley oak, coast live oak, and Oregon white oak (Quercus
garryana Hook.) have also been studied.
1
An abbreviated version of this paper was presented at the Sixth California Oak Symposium: Today’s
Challenges, Tomorrow’s Opportunities, October 9-12, 2006, Rohnert Park, California.
2
Research Forester, Forest Inventory and Analysis, Pacific Northwest and Southwest Research Stations,
Anchorage Forestry Sciences Lab, 3301 C Street Suite 200, Anchorage, AK 99503. e-mail:
tbarrett@fs.fed.us.
3
Analyst, Forest Inventory and Analysis, Pacific Northwest and Southwest Research Stations, P.O. Box
3890, Portland, OR 97208. e-mail: kwaddell@fs.fed.us.
323
GENERAL TECHNICAL REPORT PSW-GTR-217
It is impractical to monitor statewide sustainability of oak populations using age
information, as seedlings and saplings are not bored due to concerns about harming
the trees. Tyler and others (2006) suggest that the size structure of oak populations
can serve as a method for monitoring sustainability, despite the only moderate
correlation between size and age in oak populations. Diameter at breast height (DBH)
measurements have been the most frequently used size-class method for monitoring
tree populations. Numbers of trees by diameter class are used here as a direct metric
for monitoring oak species.
Assessing whether regeneration of oak woodlands is sufficient for sustainability
would require knowing rates of regeneration, growth, and mortality. With a diameter
class model for sustainability, these rates would be expressed in terms of numbers of
trees entering and leaving specific diameter classes within specific time periods. A
simple diameter class population model for a tree species can be written as:
D max
[1] N0 =
∑r N
i
i
i =1
D max
[2] Ni =
∑G N
j =0
ij
j
- miNi
∀ i = 1, .., Dmax
[3] NDmax + 1 = 0
where Ni is the number of trees in the diameter class i, Dmax is the maximum diameter
class, ri and mi are the regeneration and mortality rates for diameter class i, and G is a
vector of growth (or shrinkage) rates for transitions from each diameter class j into
diameter class i. When growth (G), mortality (m), and regeneration (r) are constant
over time, the distribution of the population over diameter classes would be expected
to slowly approach equilibrium regardless of the initial starting distribution, or, if in
the equilibrium distribution, would stay constant. For a given tree species, the shape
of an equilibrium diameter distribution of the entire population takes the form of an
inverted J-shaped curve. Species with low-regeneration and low-mortality rates can
be expected to have characteristically different diameter distributions than species
with high-regeneration and higher-mortality rates, but both will resemble an inverse
J-shaped curve. Diameter distributions that shift over time can either indicate recent
changes in mortality, growth or regeneration, or they can indicate a past change to
which the population is continuing to adjust.
The first (1981-1984) statewide inventory of oak woodlands found low
regeneration of some species (Bolsinger 1988), as did the second (1991-1994)
inventory (Waddell and Barrett 2005). In this paper, we present initial results from
the third (2001-2005) statewide inventory of oak woodlands in California, focusing
on regeneration and sapling recruitment. Diameter class distributions are presented as
a snapshot indicator of population-level dynamics.
324
Regeneration of California Oak Woodlands 2001-2005—Barrett
Methods
Data were from the annual Forest Inventory and Analysis program, which inventories
forestland in the United States (Gillespie 1999) through a combination of remote
sensing data and field plots. All forestland in California—public and private—is
monitored through this program, with one-tenth of all field plots measured each year
beginning in 2001. For this project, we used the 2001-2005 data, consisting of 8,328
plots of which 2,754 were at least partially forested. Although these data represent
only half of the plots scheduled to be installed in California, the sampling intensity is
equivalent to that used for oak woodland in the 1991-1994 inventory and roughly
twice the sampling intensity of the 1981-1984 inventory used by Bolsinger (1988) in
the first oak woodlands report.
Other changes between the second and third inventories include a change to the
procedure used to determine forestland, changes in the layout of the field plot,
changes in the procedures used to classify forest types, and a change from variable
radius to fixed radius tree selection. In addition, all forestland is included in 20012005 inventory, whereas the second inventory excluded reserved land such as state
and national parks, and the first inventory excluded reserved land and national forest
land.
Each field plot consists of four subplots, which are themselves composed of
nested circles within which trees of various sizes are measured for DBH, height,
species, and live crown ratio (USDA 2005). Trees larger than 24.0 inches DBH are
measured within a 58.9-foot radius circle (a “macroplot”), trees 5.0 to 23.9 inches
DBH are measured on a 24-foot radius circle (a “subplot”), and saplings less than 5.0
inches are measured on a 6.8-foot radius circle (a “microplot”). The total plot area on
which forest conditions are mapped is 1 acre.
All data used here are available through the national FIA Web site
(www.fia.fs.fed.us/) or by request at the regional FIA Web site
(www.fs.fed.us/pnw/fia/). Diameter distributions, tree per acre estimates, and
sampling error were calculated with standard methods (Bechtold and Patterson 2005).
Forest types were calculated using a classification algorithm (Stanford and others
2001).
Results and Discussion
Relative to other forest types, blue oak and valley oak forest have sparse seedling
occurrence (fig. 1). The metric used to measure relative abundance of seedlings is to
count the number of microplots into which at least one seedling fell. Although this
metric can be useful for comparing relative seedling abundance between forest types,
or for monitoring relative seedling abundance over time, sparse seedling occurrence
does not necessarily indicate seedling occurrence that is below replacement level.
325
GENERAL TECHNICAL REPORT PSW-GTR-217
Tanoak
Douglas-fir
California black oak
Canyon live oak / interior live oak
California mixed conifer
Red fir
Oregon white oak
White fir
Lodgepole pine
Coast live oak
Ponderosa pine
Valley oak
Blue oak
0%
10%
20%
30%
40%
50%
60%
70%
80%
90%
100%
Nonstocked Lightly Stocked Stocked
Figure 1—Percent of FIA plots 2001-2005 by seedling stocking class. Includes
seedlings of all species within plots of homogenous forest type. Nonstocked = 0 of 4
microplots with seedlings; Lightly stocked = 1 of 4 microplots with seedlings; Stocked
= 2 or more microplots with seedlings.
For the 2001-2005 inventory, the diameter distributions of interior live oak
(Quercus wislizeni A.DC.) and California black oak (Quercus kelloggii Newb.) both
resemble the classic inverse-J shape (fig. 2). California black oak’s distribution
differs from interior live oak’s distribution as would be expected of a species with
relatively lower rates of regeneration and mortality.
326
Regeneration of California Oak Woodlands 2001-2005—Barrett
350
300
Numbers of trees (millions)
250
200
California black oak
Interior live oak
150
100
50
0
1 - 2.9
3 - 4.9
5 - 6.9
7 - 8.9 9 - 10.9
11 12.9
13 14.9
15 16.9
17 18.9
19 20.9
21 +
Diameter class (in)
Figure 2—Diameter class distribution for California black oak and interior live oak, all
California forestland 2001-2005.
The common oak species that are found in more open xeric conditions have a
greater departure from typical diameter distributions (fig. 3). The distribution curve
for Oregon white oak most closely resembles typical tree species, while blue oak,
coast live oak, and valley oak all show convexity over part of their distribution. Some
of this may be explainable by sampling error, which is particularly high for trees less
than 5 inches diameter at breast height (DBH) because these trees are only measured
on the very small (6.8-foot radius) microplots.
There are several reasons why convexity in a diameter class distribution could
occur. Even with constant regeneration and mortality rates, variations in growth rates
could explain this. When mortality rates are very low, a slower growth rate in a larger
diameter class can result in convexity, as individual trees are in the small diameter
class for a shorter period of time. In addition to diameter growth changes caused by
aging, sapling growth can be increased by short-term environmental changes such as
higher levels of precipitation or seasonal changes in the distribution of rainfall
(Hanson and others 2001).
Convexity could be caused by mortality rates that are higher for certain diameter
classes for a specific interval of time. For example, a decade with larger areas of
forest burned, if the fires cause higher mortality in smaller diameter classes, could
create convexity in the population distribution. These types of causes would be most
likely to occur for species that have small geographic ranges, as the probability of a
large percent of the population being affected by a stochastic event is higher.
327
GENERAL TECHNICAL REPORT PSW-GTR-217
80
70
Number of trees (millions)
60
50
Blue oak
Coast live oak
40
Oregon white oak
Valley oak
30
20
10
0
1 - 2.9
3 - 4.9
5 - 6.9
7 - 8.9
9 - 10.9
11 12.9
13 14.9
15 16.9
17 18.9
19 20.9
21 +
Diameter class
Figure 3—Diameter distributions for blue oak, coast live oak, Oregon white oak and
valley oak, all California forestland, 2001-2005.
A third explanation for convexity in a diameter distribution is regeneration rates
that have decreased over time. One method to help understand whether changes in
rates of regeneration and mortality are affecting the population is to track the
diameter distribution over time. Looking at diameter distributions of blue oak for
three successive inventories is not conclusive, but it appears that changes in
regeneration or mortality for small saplings may have occurred in the past (fig. 4).
Even when the ratio of large trees to small trees is high, mortality and growth
rates are needed to understand sustainability. For example, in the 2001-2005
inventory, the ratio of number of blue oak trees in the 1- to 3-inch DBH class to the
number of blue oak trees in the 11- to 13-inch DBH class is 3.4:1. Past FIA
inventories provide some growth information that can be used to aid understanding of
the sustainability of this ratio. Average diameter growth for blue oak saplings
between 1981 and 1984 and 1991 and 1994 was 0.53 inches per decade for 1- to 3inch DBH trees, 0.61 inches per decade for 3- to 5-inch DBH trees, 0.60 inches per
decade for 5- to 7-inch DBH trees, 0.68 inches per decade for 7- to 9-inch DBH trees,
and 0.70 inches per decade for 9- to 11-inch DBH trees (n=37; 41; 39; 34; 37).
Assuming a 0.70 inch/decade growth rate, an individual tree might take 140 years to
grow from 2 inches DBH to 12 inches DBH. With the assumption of 140 years for
growth, the annual survival rate for 2-inch DBH trees growing to 12 inches DBH
trees would have to be 99.1 percent to sustain the 3.4:1 diameter class ratio that was
observed in the 2001-2005 inventory.
328
Regeneration of California Oak Woodlands 2001-2005—Barrett
50,000,000
1981-84
-
90,000,000
1991-94
-
90,000,000
2001-05
21
+
3.
0
1.
0
-2
.9
-4
5. .9
0
-6
7. .9
0
9. 8.9
0
11 10.
9
.0
13 12
.0 .9
15 14 .
9
.0
17 16 .
9
.0
19 18
.0 .9
-2
0.
9
0
Figure 4—Diameter distributions of blue oak from three successive forest inventories
of California. Absolute numbers are not comparable because of differences in the
included land base.
329
GENERAL TECHNICAL REPORT PSW-GTR-217
Overall, saplings are sparse in blue oak, coast live oak, and valley oak forest
types (table 1). These three oak forest types, with corresponding species that depart
furthest from the classic inverse-J diameter distribution, also have the lowest overall
density, measured in trees per acre (table 1). With open stand structures (few trees of
any size) even moderate changes in regeneration can have long-lasting effects on the
size distribution of the population. Although a low-mortality rate should contribute to
the stability of a population, a tree species that is characterized by slow growth, slow
regeneration, and a low-background mortality rate also has less potential to recover
from disturbances. Because blue oak and valley oak are endemic to California, and
coast live oak is endemic with the exception of a few small areas in Mexico, low
regeneration observed from this statewide monitoring could indicate problems with
sustainability for these three species.
Table 1—Number of trees per acre on oak woodland forest types in California, 2001 to
2005.1
Diameter class (in)
1 – 4.9
5 – 9.9
10 – 14.9 15 – 19.9
20+
All
classes
Mean
Mean
Mean
Mean
Mean
Mean
Forest type
(SE)
(SE)
(SE)
(SE)
(SE)
(SE)
Blue oak
60.5
49.2
15.9
5.2
2.3
133.1
(10.2)
(2.6)
(0.9)
(0.5)
(0.2)
(11.1)
Canyon live –
345.2
109.7
26.3
7.4
4.6
493.1
interior live oak
(27.8)
(5.4)
(1.5)
(0.6)
(0.4)
(28.3)
Coast live oak
77.8
58.5
28.4
14.7
7.2
186.5
(18.7)
(5.1)
(2.8)
(1.4)
(0.8)
(20.2)
California black oak
242.7
90.2
29.6
10.5
6.5
379.5
(31.4)
(7.2)
(2.1)
(1.0)
(0.6)
(33.8)
Valley oak
22.2
45.8
13.0
2.3
4.8
88.0
(9.8)
(10.3)
(3.1)
(0.7)
(1.3)
(15.2)
Oregon white oak
227.8
91.9
24.0
7.2
2.8
353.7
(35.2)
(8.7)
(2.3)
(1.2)
(0.7)
(36.4)
1
Includes trees of all species
Conclusion
Based on the 2001-2005 inventory, sapling recruitment of blue oak, valley oak, and
coast live oak appears to be low for the statewide population. These species show
diameter distributions that are atypical for forest tree species, but these species are
also atypical in their very open low-density woodland structure. Sparse regeneration
and low density of small saplings for blue oak has been consistently observed for
three decades of forest inventory, and in observational records for nearly a century.
Diameter distributions are useful in understanding the populations’ structures, but it
is not possible to know whether the lack of regeneration results in non-sustainable
populations from diameter distributions alone. Better information on statewide
regeneration (ri), growth (Gij), and mortality rates (mi) would allow dynamic
modeling of the population structure and enhanced ability to monitor for long-term
sustainability for these species.
330
Regeneration of California Oak Woodlands 2001-2005—Barrett
References
Arner, S.L; Woudenberg, S.; Waters, S.; Vissage, J.; MacLean, C.; Thompson, T.; Hansen, M.
2001. National Algorithms for Determining Stocking Class, Stand Size Class, and
Forest Type for Forest Inventory and Analysis Plots. Internal document. On file with
the Forest Inventory and Analysis program, Portland, OR: U.S. Department of
Agriculture, Forest Service, Pacific Northwest Forest and Range Experiment Station.
Bechtold, W.A.; Patterson, P.L., eds. 2005. The enhanced forest inventory and analysis
program – national sampling design and estimation procedures. Asheville, NC:
USDA Forest Service General Technical Report SRS-80. 85 p.
Bolsinger, C.L. 1988. The hardwoods of California’s timberlands, woodlands, and
savannas. Resour. Bull. PNW-RB-148. Portland, OR: U.S. Department of Agriculture,
Forest Service, Pacific Northwest Research Station. 148 p.
Gillespie, A.J.R. 1999. Rationale for a national annual forest inventory program. Journal
of Forestry 97:12:16-21.
Hanson, P.J.; Todd, D.E., Jr.; Amthor J.S. 2001. A six-year study of sapling and large-tree
growth and mortality responses to natural and induced variability in precipitation
and throughfall. Tree physiology 21:345-358.
McDonald, P.M. 1990. Quercus douglasii Hook. & Arn. – Blue oak. In Burns, R.M. and
Honkala, B.H., Tech. Coord., Silvics of North America, Vol 2, Hardwoods. Washington,
DC: USDA Forest Service Agricultural Handbook 654. 877 p.
Rizzo, D.M.; Garbelotto, M.; Davidson, J.M.; Slaughter, G.W.; Koike, S.T. 2002.
Phytophthora ramorum as the cause of extensive mortality of Quercus species and
Lithocarpus densiflorus in California. Plant disease 86:3:205-214.
Sudworth, G.B. 1908. Forest trees of the Pacific slope. Washington, DC: U.S. Department
of Agriculture, Forest Service. 441 p.
Tyler, C.M.; Kuhn, B.; Davis, F.W. 2006. Demography and recruitment limitations of
three oak species in California. The quarterly review of biology 81:2:127-152.
Tyler, C.M.; Mahall B.E.; Davis, F.W. 2002. Factors limiting recruitment in valley and
coast live oak. USDA Forest Service Gen. Tech. Rep. PSW-GTR-184. p. 565-572.
United States Department of Agriculture [USDA]. 2005. Field instructions for the annual
inventory of Washington, Oregon, and California. Portland, OR: Forest Inventory
and Analysis, Pacific Northwest Research Station. www.fs.fed.us/pnw/fia.
Waddell, K.L.; Barrett, T.M. 2005. Oak woodlands and other hardwood forests of
California, 1990s. USDA Forest Service Resource Bulletin 245. Portland, OR: Pacific
Northwest Research Station.
Continue
331