SEED BANKS OF LONG-UNBURNED STANDS OF MARITIME CHAPARRAL: COMPOSITION, GERMINATION BEHAVIOR, AND SURVIVAL WITH FIRE

9 Pages·2000·7.9 MB·English

by  Dennis C Odion

Madrono, Vol. 47, No. 3, pp. 195-203, 2000
SEED BANKS OF LONG-UNBURNED STANDS OF
MARITIME CHAPARRAL: COMPOSITION, GERMINATION BEHAVIOR,
AND SURVIVAL WITH FIRE
Dennis C. Odion'
Department of Geography, University of California, Santa Barbara, CA 93106
Abstract
Seed germination requirements in the California chaparral have been described mainly from freshly
collected seed. However, uncertaintiesremain becausethebehaviorofseeds in the soil candiffer. Istudied
germination ofthe seed bank in long-unburned stands ofmaritime chaparral in central coastal California.
I quantified seedlings emerging from soil samples provided with appropriate temperature and moisture
conditions following I) no othertreatment, 2) a heat treatment to optimize germination ofheat-stimulated
species, 3) the same heat with the addition ofcharred wood, and 4) the burning ofchaparral stands prior
to collection ofsamples. I compared germination in these treatments with seedling emergence in the field
following fire. I also collected and divided samples into 0-2.5 and 2.5-7.5 cm depth fractions to evaluate
abundance of seed at the surface and depth before and after fire.
Seed of one annual had reduced germination following the heat treatment. Seeds of all other species
common enough to evaluate statistically were heat tolerant. However, because seeds were found to be
mostly near the surface, there was considerable mortality with fire. Moreover, seedling populations in the
field only accounted for a fraction of the seed bank that survived fire, and seventeen species that ger-
minated in samplesdidnotgerminate and/oremergeinthefield. Mostspecies' germinationandemergence
was influenced in some way by heat and/or charate. Germination of two Ceanothus was dependent on
heat. Adeuostoma fascicidatum Hook. & Arn., Arctostaphylos piirissima P. Wells, and two annuals had
germination that was enhanced by heat and enhanced further when charate was added. Despite the im-
portance of fire effects, there were no short-lived species having entirely fire-dependent germination.
Germination and/or emergence of 3 species was negatively affected by charate. These germinated spar-
ingly or not at all after fire.
One ofthe most prominent evolutionary special- tion without fire may also be inhibited by allelo-
izations to fire exists in the germination ecology of pathic chemicals washed from foliage or litter
seeds from plants found in Mediterranean shrub- (Muller et al. 1968; McPherson and Muller 1969),
lands, particularly those ofAustralia, South Africa, and/or phytotoxins produced by soil microbes (Ka-
and California (Bond and Van Wilgen 1996). This minsky 1981). Fire eliminates these compounds.
subject has received considerable attention (Review Seeds ofchaparral plants range from readily ger-
by Keeley 1991), revealing a fascinating complex- minable at the time of dispersal (non-refractory) to
ity offeatures that insure germination will coincide deeply dormant (refractory) as a result of multiple
with the anomalously favorable conditions forseed- barriers to germination (Keeley and Fotheringham
ling establishment that exist after fire. There are 1998). Some species produce a portion of seed that
physical features such as bradyspory (or serotiny) is refractory and a portion that is not (Emery 1988;
where seeds stored in fruits and cones are released Parker and Kelly 1989). Generalizations about the
when heated by fire (Whelan 1995), and impervi- type(s) of dormancy species exhibit derive mainly
ous seed coats that open with the heat offire (Swee- from tests on freshly collected and stored seed.
ney 1956; Quick and Quick 1961; Auld and Germination ofseeds residing in the soil may differ
O'Connell 1991). Physiologically dormant seed
cmaalys wbaesihnedducferdomtocghrarorwedfowllooowding(Wfiirceklboywch1e9m7i7-; tsiogmnaififcaasnctilyc,idaastuhmasHoboeke.n &docArunm.endtineddAfrocrtoAsdteaupohsy--
los canescens Eastw. among others (Stone and
Keeley 1984, 1987; Keeley et al. 1985; Keeley and
Juhren 1951; Parker 1987; Keeley and Fotheringh-
Pizzorno 1986), water soluble nitrogenous com-
pounds (Thanos and Rundel 1996) and smoke am 1998). Seeds exposed to allelopathic chemicals
(Keith 1997; Keeley and Fotheringham 1997, and phytotoxins found in chaparral soils may ex-
1998). For each fire-related germination cue, there hibit enforced dormancy (Muller et al. 1968; Mc-
are multiple dormancy-releasing mechanisms that Pherson and Muller 1969; Keeley 1991). Therefore,
have evolved convergently among disparate floras it is imperative to study the soil seed bank to un-
(Baskin and Baskin 1998). In chaparral, germina- derstand how chaparral germination is controlled in
nature.
The potential for germination in the chaparral
' Present address: Marine Science Institute, University seed bank without fire is thought to be low for most
ofCalifornia, Santa Barbara, CA 93106. species because seedlings are rarely apparent under
MADRONO
196 [Vol. 47
the shrub canopy. However, Christensen and Muller lish that the chaparral had not burned for 75-80 y
(1975), Tyler (1995), and Swank and Oechel (1991) at the site where more intensive sampling was un-
reported considerable seedling growth under Ad- dertaken (site 1). Samples were also collected from
enostoma in plots protected from herbivory, sug- a second, nearby site which had not previously
gesting more germination can occur without fire burned for about 50 y. Both sites were dominated
than is evident. In addition, Zammit and Zedler by Adenostoma.
(1988, 1994) and Holl et al. (2000) found that many
species germinated readily from chaparral soil seed Methods
bank samples. To test how much germination can
occur in species' seed banks without fire and how Transects consisting of 47 contiguous 1 m^ plots
much requires heat and/or chemicals produced by were established in dense chaparral dominated by
fire, I compared emergence from controls and uni- Adenostoma. Nine 5 cm diam, 7.5 cm deep cores
form fire treatments that are known to maximize of soil were obtained per plot at site 1 in the fall
germination of refractory seed without inducing of 1988. Five cores were collected in fall, 1989 at
mortality. I then analyzed how much of the in situ site 2. Chaparral at both sites was burned soon
seed bank was eliminated above and below 2.5 cm thereafter, with low fuel moisture contributing to
in the soil by fire in the chaparral stands. Finally, I relatively intense fires (Odion and Davis 2000). I
enumerated seedling emergence in the field to com- collected 5 cores per plot the day after each fire.
pare germination in nature vs. in collected samples. Seed bank cores foreach plot were composited, and
The chaparral I studied is geographically-isolated 350 cc subsamples were removed from each ho-
and its environment differs in many respects from mogenized sample.
that found in the Transverse and Peninsular ranges Pre-burn samples from site 1 were given three
inland. With Santa Ana winds absent, and a lower treatments: 1) heat, 2) heat and charate (charred,
frequency of ignitions, coastal environments have pencil-sized Adenostoma stems collected after the
likely supported less frequent and dynamic fire, at fire and ground up) and 3) control (no heat or char-
least prior to human dominance of the fire regime ate). Only the second of these treatments was used
(Odion et al. 1992; Odion et al. 1993). The average on pre-burn samples from site 2. Samples to be
mm
lifespans of the Ceanothus spp. in maritime chap- heated were spread to a depth of 2-3 on alu-
arral are particularly short (Davis et al. 1988) com- minum cooking trays. Based on studies by Wright
pared to those inland (Keeley 1975, 1992). Death (1931), Sampson (1944), Sweeney (1956), Keeley
ofthe non-sprouters opens space for recruitment by and co-workers (several publications, see Keeley
numerous herbs and subshrubs (Odion and Davis (1991), heating at 100°C for ~5 min typically pro-
2000). These and other factors such as soil and cli- duces the greatest germination response among fire-
mate may help explain differences in post-fire re- recruiters, and is well within their heat tolerance.
generation in maritime vs. nearby inland chaparral Given the slight insulation the soil would provide,
reported by Tyler (1995); they may also have con- I decided to use a 7 min duration. Heat-induced
tributed to evolutionary divergence in maritime seed mortality is controlled predominantly by max-
chaparral taxa that has produced endemic Arctos- imum temperature, as opposed to duration (Borch-
taphylos and Ceanothus (Griffin 1978). I have eval- ert and Odion 1995), so it is unlikely that this
uated my germination data for any evidence that change effected mortality. Heating was done in a
the environment and insularnature ofthe study area forced-air oven.
manifested variation in germination ecology. The subsamples were spread on sterile sand in
20 cm plastic pots. The amount of charate added
Study Area was 2 rounded tablespoons (21.3 ± 0.94 g, n =
11).
Samples were collected fromAdenostomafascic- The pots were covered with clear plastic, pro-
ulatum (hereafter Adenostoma) chaparral located tected from herbivory, and kept moist out-of-doors
near sea-level, within Vandenberg Air Force Base under 50 percent shade cloth at Cal-Orchid Nursery
in central, coastal California as described in in Santa Barbara, where temperature fluctuations
D'Antonio et al. (1993), and Odion and Davis were analogous to the field. Potting was complete
(2000). Substratum here is Pleistocene eolian sand in late November, at which time all samples were
(Dibblee 1950). Climate in the area is strongly in- given their first watering. All the samples were ex-
fluenced by the prevailing onshore winds and cool posed to outdoortemperatures from the time ofcol-
ocean, and the temperature regime is mild, es- lection through the subsequent growing season to
pecially for a chaparral environment. Maritime provide natural temperature stratification.
chaparral of the area has been described in detail Seed bank sampling was also undertaken at ran-
m
by Davis et al. (1988). The average annual precip- domly located --1.5 diam canopy gap areas ad-
itation is 35.3 cm. I counted annual rings from the jacent to the site 1 transect. I took samples from
obligate seeders, Arctostaphylospurissima P. Wells, the center of the gap as well as the edge and un-
and Ceanothus cuneatus (Hook.) Nutt. var. fasci- derstory ofthe adjacentAdenostoma canopy. These
cularis (McMinn) Hoover (Keeley 1993) to estab- cores were separated into 0-2.5 cm and 2.5-7.5 cm
2000] ODION: GERMINATION OF MARITIME CHAPARRAL SEED BANK 197
depth fractions and given the heat and charate treat- perennial Gnaphalinms, had germination that was
ment. Due to smaller amounts available, 175 cc not affected by the heat treatment (Table 1). One
subsamples were spread over sand in 16 cm diam of these, G. microcephalum Nutt. was significantly
plastic pots placed with the others. negatively affected by charate. Among heat-affect-
All germinants were identified and removed from ed species, the two obligate-seeding species of Ce-
pots through the growing season. Nomenclature anothus, the subshrub Helianthemum scoparium,
follows Hickman (1993). Specimens whose identity and the annual Thfolium microcephalum Pursh had
'I was uncertain were grown until it was determined. significantly greater germination with heat alone,
Five of the pre-burn samples treated to heat and while the opposite occurred for the annual Calan-
I charate were removed from pots after emergence drinia ciliata (Ruiz Lopez & Pavon) DC. (Table 1).
stopped. I repotted these the following autumn. No Other important species that had a positive response
further germination occurred in these. to heat were also affected by charate. Adenostoma,
Arctostaphylos purissima, and Lotus strigosus had
Results significantly greater germination with heat and
charate compared to heat alone.
General patterns. Seed from 72 species germi- Germination of Centaurium davyii and Crassula
nated and emerged from samples collected along connata with heat and charate was not only signif-
transects (Table 1). More than half (48) were an- icantly lower than with heat alone, but also lower
nuals. Site 2 had greater diversity (60 vs. 48 spe- than with no treatment. Crassula was rare in the
cies). Many of the same species emerged abun- burn areas, and Centaurium did not occur there un-
dantly from samples from both sites (e.g., Aden- til the third year after fire. Both species were fairly
ostoma, Helianthemum scopariiim Nutt., Crassula common in the surrounding unburned chaparral.
connata (Ruiz Lopez & Pavon) A. Berger, Cen- Mimulusfloribundus had much lower germination
taurium davyi (Jepson) Abrams, and Navarretia with heat and charate than with heat alone, but heat
atractyloides (Benth) Hook & Arn.). Despite this, and charate germinants outnumbered those in con-
there were only 33 species in common. In addition, trol samples ( P > 0.05, NS). This species, though
two subshrubs, Miniidus aiirantiacus Curtis and Lo- abundant in several samples, was absent from most.
tus scoparius (Nutt.) Ottley, were abundant in site It was never observed in the field, including in un-
2 samples and absent in those from site 1. Several burned chaparral. It is typically found in seasonal
annuals were common in samples from one site but wetlands like two other species that were found in
not the other. Nine species were non-native. All 9 samples, but not in the field, Crassula aquatica (L.)
are widespread weeds. Schonl. and Centunculus minimus L.
Post-burn samples contained 44 species (1 non- With a relatively high proportion ofseed at depth
native) and substantially reduced numbers of ger- (76 percent below 2.5 cm in gap, edge, and under-
minants. The reduction varied with species depend- story samples combined. Table 2), Arctostaphylos
ing on the proportion of seed present at depth, as purissima had better survival (post-burn/pre-burn =
described below. Reduced germination was also 17 percent) than Adenostoma (site 1 = 2 percent,
strongly correlated with the amount of soil heating site 3 = 3 percent) which only had 22 percent of
that occurred where samples were located (Odion its seed below 2.5 cm. These survival percentages
and Davis 2000). Thus, the number of post-burn are from transect samples. The 2.5-7.5 cm depth
germinants was much greater in samples from gaps samples had relatively little emergence of species
vs. under the shrub canopy. Horizontal patterns of with charate-enhanced germination. Seeds of other
seed abundance are analyzed in Odion and Davis shrubs were not abundant enough to evaluate depth
(2000). distribution. The high survival of Ceanothus cu-
There were 20 and 31 species that germinated in neatus at site 2 (33 percent) as well as results from
the field respectively at the two sites (Table 1). a fuel translocation experiment (Odion and Davis
Twelve were at both. Four species germinated in 2000) suggest this obligate-seeder had a high pro-
the field plots but not seed bank samples. There was portion of seed at depth.
only one individual ofeach. Seventeen species ger- Helianthemum scoparium was particularly abun-
minated in samples but did not germinate and/or dant and not affected by charate, so the effect of
emerge in the field, including species whose seed the depth distribution of its seed is relatively clear.
was among the most abundant (e.g., Centaiiriiim The subshrub had 53 and 75 percent of its seed
davyi, Mimulus floribiindiis Lindley). Another, bank below 2.5 cm in gap and understory samples
Crassula connata, was virtually absent in the field respectively. Despite the greater proportion of seed
in the burn areas, although it was common in ad- at depth at understory plots, survival was similar
jacent unburned chaparral. Field populations for there (22 percent) compared to gaps (24 percent).
most species were mu—ch smaller than post-burn Survival along the site 1 transect was 9 percent
seed bank populations between ~5 and 14 times (post-burn/pre-burn heat). After fire, 94 percent of
smaller for shrubs, and generally even smaller for Helianthemum scoparium seeds were below 2.5 cm
other species. in the soil in understory plots, only half were at
Germination treatments. Only two species, both depth in gap plots.
MADRONO
198 [Vol. 47
tu c
C3 O' d(N o o d O" o O" o mO5 O q" qin '^i; rn O q q 0^0' o o o oo o
C/5
1^
xj; \—qI -vt; ^^ ^ (qN -xt; oov^IdT)r(^Nn(^rN^m00doo OO^ C0N0 O ^' O ^ q(N T^j- Ti- I
m V
„1 ^ 0^^00^^^)000m'^ Oo0qd0o0^INT0)o0d0rm00j00u0r--0-i0o0d^^ O 0i0n ^ o o r00nOr'n-OHNcrn-i^n^rc-N^-i^Ooi^noNCON(NOO
o<^Nr(n^—;d^-0^0doooo ^do^oooooo o o o o o o o ooO'(-NH"-HodoooOdoo
5< -O ^SC3
'^H
^< s2 s3 or-^odoO(^oooo ooooooooo ^' O O (N O O ^ ooooooo(N"0io0n od00o0d0o
P3 cd 'o
ootoodo00d00dfrN^oooo OOO^OOOOO 0000 o mino^ino d00 o o^or-<^Nvdooin(Nr-^'m0d0o00do
»n in<N \D
i-L- cad - r-H (N
zee
c/) <U
2 o^^do0^0o0i0or(*o-N)Nr*-') ^xt-ooo rl<-)OO0O0 ^^OOOO (\N0 (-NH O0q0nm^'xt; O^^ ((NN ooooNOsmodoiinnr"v^O ^-^r-d^
z - 2
UJ ^ (U
Q< ^<u SC OO^Or^oOO((NNOOO r~^o^ooooo ^X)- ^^ o0d0 'rv-O' o 0000 ^^ ooood00or^-^H ^or-^i^^n<(NNm00doo
aX- ^ oj
7 (/2
is 5 tso
"a
C3 O 5
IP
•S
oo ^R Q^ 1-5 II
z + III Q ^s oS Ka ^ 0 ^ a a
ll-el a ^ o s 'S
D3 ^ §111 00 a a < « ^ K -
< O GO ~ t; ^ ? .U'^ 'i: ^ ^ kJ ^ U!a^ Oas Oas: <Z C C ^Ci,^Ci,UQ u u o u u
ODION: GERMINATION OF MARITIME CHAPARRAL SEED BANK 199
oo (oN'-inh'o Oon'oooooo o o ooooooooo in o
r-^0'^'r^oooc»o^'oo^r~^o<N(^o^oooNo6o6oooooooooorj o o
o
O0f—n0^N—^'"^^^^r'"'-^nrOxi^^^^r^^->^^or'c^^N^^oo^rN~.-''x^t^'-r^r-^nHroo(i0oNn0o0o0 oi(^^Noo-HM(N3'(iKNn(riNn^O''^^^ir^n-^o^''^ooo°o96o^^^0ro0njo
d(Nooooooodoooooodor-H<Nin—m o«noooooooooooooooo
'o
^ ^ ^
ooooooooooooooo-^ or<Nnr^vDoodoooooooooooooooo00o
—
1 CM 00
^'^(N
OOiONnOoO-6Vo^Or-~"oo^o<^No»rn^ooo6O'O^'"^-^r^C-Nr^^^OVNmDaiNnrr~n0^0oO'^ooooO'xf'N^oooooo-O'^^r^'0"^0^o
^
00 r-- --^
O(mNN) o o ir(-Nn" o o ^^' ^'vt;(0iN0n Oinoo0d0—\^D^^~K^(rNi-nO'rOr)d^ion(oN^rno^^<N ooooooooooor-^o(Noo0'0 0^0o
CO
oooq^tTtTtoocNooTtoq^MD
^^.'oO'^^.'ooo6odvo^"-^'ood(^o6o6o6rt-r-^oorNioo^ooooo^'^o.rn'xt-oo^^
o ^ m
l>5
2 s
2 ^ 8 g
I h
is liI § CO
Q
1-2 ::::
o o o "~5 2 ^ t~ S ^to Co ^
5 ^ t>o So i: cx,>- cs^ 5 ^^ --^ ^ Scioo s- ^ ^
MADRONO
200 [Vol. 47
Table 2. Numbers of Germinants, Expressed as Average Number per m% From 0-2.5 cm and 2.5-7.5 cm Depth
Fractions Before and After Fire, and in the Field From the Same Plots in Which the Separate Depth Samples
Were Taken. Values are the averages from 30 samples expressed as the density ofseed per m-. Pre-burn samples were
treated with heat and charate.
pre-burn post-burn field
0-2.5 2.5-7.5 0-2.5 2.5-7.5
Shrubs
Adenostomafasciciilatiun Z4U.1 jZ.Vz OQ.OO 5.4
Arctostaphylospurissima 44.1 141.1 36.8 26.5 10.6
Subshrubs
Helianthemiim scoparium 411.6 582.6 88.2 145.6 47.0
Perennial herbs
Mimulusflorihundus 1519.0 538.0 39.2 26.5 0
,
Annual herbs
Apiastrum angustifoliiim 245.0 26.5 0 17.6 0.8
Centaurium davyi 1396.5 608.6 558.6 185.2 0
Crassula connata 6056.4 1525.9 1 166.2 493.9 ' 1.6
Cryptantha clevelandii 78.4 17.6 0 0 0.1
Lotus strigosus 83.3 88.2 19.6 79.4 11.0
Navarettia atractyloides 3013.5 299.9 274.4 61.7 29.7
Survival percentages ofthe seed bank for species mination in the absence of fire for this to occur.
whose germination was negatively affected by Consistent with this, I found that most of the seed
charate are equivocal because post-burn samples bank formany species needed fire to germinate (Ta-
presumably contained the inhibitors(s). Among the ble 1). Two shrubs, both species of Ceanothus, had
remaining annuals, high mortality was common. In germination that was entirely fire-dependent. How-
fact, Apiastrum angustifoliwn Nutt. though fairly ever, I also found that a small but distinct portion
common in pre-burn samples, was not detected in of the seed bank for all other fire-recruiters ger-
post-burn transect samples, and was rare in the minated with simply moisture and natural temper-
field. Seed of this diminutive plant was predomi- ature fluctuations. In addition, there were no short-
nantly near the surface (Table 2). The second most lived species detected that had entirely fire-depen-
abundant species in the field after fire was the an- dent germination {i.e., there were no specialized fire
nual Navarretia atractyloides. Combining data annuals), which is not typical for chaparral. Seed-
from gap and understory plots in Table 2, while lings of these are usually found only the first year
better illustrating survival at depth, obscured other after fire (Keeley et al. 1981). Short-lived species
patterns. Where seed of this species was concen- in my study all produced seedlings after the first
trated, in canopy gap areas, only 6 percent of its post-fire growing season (Odion 1995, unpublished
seed was below 2.5 cm in the soil, explaining why data). Thus, fire-recruiters in this study, other than
only 10 percent survived there despite relatively Ceanothus, produce seed that is both refractory and
low soil heating with fire. Conversely, in the un- not. Based on my germination results and those by
derstory, 16 percent of seed was in the deeper frac- Davis et al. (1989) and Holl et al. (2000), as well
tion, explaining the relatively high survival (22 per- as extensive field observation (Davis et al. 1988;
cent) along the site 1 transect (predominantly un- Odion et al. 1992; Odion et al. 1993) non-refractory
derstory). Seed ofLotus strigosus (Nutt.) E. Greene seed may be somewhat more important in maritime
was equally abundant in deep and shallow samples vs. inland chaparral, at least among short-lived spe-
overall (Table 2). Survival in surface samples was cies. Conversely, forAdenostoma, the proportion of
a relatively high 24 percent, and at depth 90 per- non-refractory seed (19%, Table 1) is in agreement
cent. with what has been found at more inland locations
(Stone and Juhren 1953; Zammit and Zedler 1988).
Discussion How might having seed that is both refractory
My procedures indicated that seed mortality with and not be a selective advantage in chaparral? By
fire was substantial, and greater in the older stand. producing seed that germinates readily in the field,
Previous studies have also found that a significant short-lived species may grow and reproduce
number of seeds do not survive fire in chaparral throughout the fire cycle, which may be critical for
(Keeley 1977; Davey 1982; Bullock 1982; Zammit them to sustain seed populations from one fire to
and Zedler 1988; Davis et al. 1989). For species to the next (Zammit and Zedler 1988). As fire interval
ensure successful recruitment after fire, their seeds increases, the capability to augment the seed bank
must accumulate at depths in the soil where they between fires will be of increasing importance to
will be safe. There must be strong resistance to ger- short-lived species because their seed banks will
2000] ODION: GERMINATION OF MA] HME CHAPARRAL SEED BANK 201
i
I
otherwise diminish due to mortality and predation. ulants can be produced when soil or wood are heat-
Therefore, considering the past likelihood of rela- ed to 175°C for 10-30 min (Keeley and Nitzberg
tively long fire intervals in maritime chaparral, it is 1984; Keeley and Pizzorno 1986). However, these
not surprising that I found non-refractory seed to stimulants are effective in very low concentrations
be so prevalent, even among fire-recruiters. It is (Keeley and Pizzorno 1986) and my heat treatment
possible that some non-refractory seed remains dor- did not produce a germination effect comparable to
mant under the chaparral canopy due to inhibitors. that found with heat and charate.
However, short-lived species such as Navarettia In contrast to Adenostoma, the Arctostaphylos
atractyloides, Helianthemum scoparium, Chorizan- and Ceanothus'' in this study are narrowly distrib-
the spp. and Camissonia micrantha (Sprengel) Ra- uted taxa whose germination ecology has not been
ven are common in old age class maritime chap- previously studied. However, congeneric ecological
arral (Davis et al. 1988; Holl et al. 2000), particu- analogs can be compared. I found that heat was
larly in the canopy gaps that typify this vegetation. effective in inducing Arctostaphylos purissima
Their seeds are concentrated in gaps, an advantage seeds to germinate, and again that there was a syn-
because survival with fire is much greater there ergistic effect with both heat and charate. This ef-
(Odion and Davis 2000). The germination ecology fect has been found in one oihQVArctostaphylosthat
of these and other short-lived species allows them coincidentally is also a narrow endemic from mar-
to exploit gaps when they appear in maritime chap- itime chaparral, A. morroensis Wiesl. & B. Schrei-
arral, resulting in more abundant post-fire recruit- ber (Tyler et al. 1998; Tyler et al. 2000). Their
ment than would otherwise occur. Such opportunis- methods avoided potential influence of soil-derived
tic germination has been documented in otherchap- stimulants because seeds were extracted from the
arral (Zammit and Zedler 1988), and linked to can- soil priorto heating. Germination doubled with heat
opy gaps (Shmida and Whittaker 1981), but its and charate, but, there was no effect with heat
relative importance undoubtedly varies with chap- alone, or charate alone. Conversely, Parker (1987)
arral canopy dynamics. found that dormancy ofA. canescens seeds extract-
The dominant shrubs in the study, Adenostoma, ed from the soil was overcome by charred wood
Arctostaphylos, and the two species of Ceanothus, extract alone. Freshly collected seed remained dor-
are fire-recruiters. Nearly all chaparral areas are mant with the same treatment. Other research using
dominated by some combination of these genera freshly collected seed has found that just heat
and there has been much interest in their germina- (Sampson 1944; Berg 1974) or charate (Keeley
tion ecology. ForAdenostoma, the most widespread 1987) can be effective in breaking seed dormancy
and abundant chaparral shrub, a question that has ofArctostaphylos. Furtherresearch onArctostaphy-
persisted had been, what is the role of heat in ger- los spp. seed banks is needed to determine how
mination of refractory seed? Adenostoma seed variation in their germination may be correlated
banks have been studied previously by Christensen with fire regime or other environmental variables.
and MuUer (1975) who heated soil under shrubs in Ceanothus spp. have a hard seed coat that can
situ Zammit and Zedler (1988, 1994) who burned be cracked by heat (Quick 1935; Quick and Quick
straw over soil placed in flats, and Parker (1987) 1961). There can be some germination in the ab-
who oven-heated soil and supplied charred wood sence of fire if the impermeability of the seed coat
extract. The first two procedures enhanced germi- deteriorates over time (Quick and Quick 1961),
nation ofAdenostoma, but it is unclear whetherthis e.g., C. greggi A. Gray, (Moreno and Oechel 1991;
was a direct or indirect effect ofheat. Parker (1987) Zammit and Zedler 1994). In addition, Keeley
found that charred wood extract, not heat, enhanced (1991) reports that it is typical for a few percent of
germination. It is possible that the heat he supplied the seeds of Ceanothus to be non-refractory. In the
(100°C for 1 h) was in excess of what the seeds in stands I studied, seeds of C. cuneatus vanfascicu-
the samples could tolerate since it resulted in a de- laris and C. impressus had resided in the soil for a
crease in germination. On seeds collected from considerable length of time. Neither species was in
shrubs, oven-heating stimulated germination the pre-burn vegetation in the older stand; both
(Wright 1931; Sampson 1944; Stone and Juhren drop out as stands of maritime chaparral age-after
1953). In addition, Keeley (1987) found that heat only —20 y in the case of C. impressus (Davis et
alone increased germination compared to controls al. 1988). Nonetheless, I found no germination of
in 6 of6 different temperature treatments, but there this species or C. cuneatus var. fascicularis without
was not a statistically significant effect. However, heat. Because input into the seed bank for these
Keeley did find significantly enhancedAdenostoma species will cease in long-unburned stands, seeds
germination with charate, and that there was a syn- must survive and remain dormant for their seed
ergistic effect with heat and charate. I also found banks to persist. These two species may have seed
that heat and charate produced a synergistic effect, coats that are especially resistant to deterioration,
but unlike Keeley, that germination was signifi- perhaps because they are relatively thick. Thickness
cantly enhanced with heat alone. It is possible that of seed coats is correlated with heat endurance
stimulatory substances were formed and/or inhibi- (Wright 1931), and I found these Ceanothus spp.
tors destroyed when I heated soils. Chemical stim- were more capable ofgerminating in areas ofgreat-
MADRONO
202 [Vol. 47
er soil heating than any other species at the two Haller, and Nancy Vivrette provided insightful feedback L
burn sites (Odion and Davis 2000). over the course ofthe study. Pat Shafroth, Joe Mailander,
Another hard-seeded species with heat-stimulat- and Kelly Cayocca assisted with the collection and treat-
ed germination was the annual Lotus strigosus (Ta- ment of seed bank samples. Clifton Smith identified a
ble 1). It is curious that this species, unlike the Ce- fnourmtbheerseocfonstpreicbiuetsi.onIs.express my sincere thanks to each
anothus spp., had germination further enhanced by
heat and charate. Lotus strigosus was common the
second spring after fire. Second year plants may Literature Cited
have emerged from seed that did not germinate the AuLD, T. D. AND M. A. O'CoNNELL. 1991. Prcdicfing pat-
first year, or from seed produced the first year. terns ofpost-fire germination in 35 easternAustralian
Charate-induced germination could allow newly Fabaceae. Australian Journal ofEcology 16:53-70.
i
pwirtohdouuctedheasteeidf wtaotegrersmoilnuabtlee btyhperofdoulcltoswionfgwyoeoadr Baskibni,ogeC.ogrCa.phayn,daJ.ndM.evoBlaustkiionn.o1f99d8o.rmSaeendcsy: aencodloggeyr,- I
combustion are still present in the burn area. An- Berg,minAa.tioRn.. A1c9a7d4.emiArcctPoresstsa,phSyalnosDiAedgaon,sC.A.manzanita. I
other hard-seeded annual legume Trifolium micro- Seeds of woody plants in the United States, USDA j
cephalum did not have charate-enhanced germina- Forest Service. !
tion, and the phenomenon has not been reported Bond, W. J. andB. W. vanWilgen. 1996. Fireandplants.
among other hard-seeded species (Baskin and Bas- Chapman and Hall, New York, New York.
kin 1998; Table 10.6). Borchert, M. I. and D. C. Odion. 1995. Fire intensity
Baskin and Baskin (1998; Tables 10.4 and 10.7) and vegetation recovery in chaparral: areview, p91-
list a few chaparral species that may have germi- 100. in J. E. Keeley, and T. Scott (eds.), Brushfiresin i
nfaotuinodngreerdmuicneadtiboyn thheaattwoarscshuaprprredeswseododbyehxteraatct(so.neI CnaaltiifoonranliaA:sseoccoilaotgiyonaonfdWrielsdoluarncde Fmiarne,agFeaimrefnietl.d,IWntAer.- j
Bullock, S. H. 1982. ReproducfiveEcologyofCeanothus
species), and by heat and charate (three species; cordiilatiis. M. A. Thesis, California State University, \
Table 1). Curiously, two of the species suppressed Fresno. !
\
by heat and charate (Crassula connata and Cen- Christensen, N. L. and C. H. Muller. 1975. Effects of
taurium davyi) had germination that was enhanced fire on factors controlling plant growth in Adenosto-
by heat alone. Both were uncommon in the field ma chaparral. Ecological Monographs 45:29-55.
after fire despite having abundant seed in the soil. D'Antonio, C. M., D. C. Odion and C. M. Tyler. 1993.
In fact, it was not until three to four years after Invasion ofmaritime chaparral by the alien succulent
each burn that Centaurium seedlings appeared, so Carpobrotiis edulis: the roles of fire and herbivory.
Oecologia 95:14-21.
tshaempslaemseapmpeacrehnatnliysompetrhaatteidnihnibtihteedfiegledr.mTihneatrieosnultisn Davesyi,folJi.usR.. 1M9.8S2.. STthaensdisR,epClaalcifeomrennitaiSntCateeanPootlhyutsecchrnaisc-
for Crassula may be at odds with what occurs else- University, Pomona.
where. This species is often apparent after fire in Davis, F W., D. E. HicksonandD. C. Odion. 1988. Com-
chaparral, however, this may be due to increased position of maritime chaparral related to fire history
biomass of individuals, not increased densities. and soil. Burton Mesa, California. Madrono 35:169- i}
In conclusion, seed banks in the maritime chap- 195.
arral I studied may differ from most inland coun- , M. I. Borchert and D. C. Odion. 1989. Estab-
ttearnpcaertosfninon-trheefrfaocltloorwyinsgeedw,ay2s):la1c)kogfreeantteirreliympfoirre-- Dibblclehieas,phamTrernatWl.oaff1tm9ei5rc0r.fiorGse.ceaoVlleeoggevtyeagtoeitfoasto8iu4otn:h5pw3ae-ts6tte7errnninSamnatraitBiamr-e
dependent germination in short-lived species, 3) bara County. Bulletin 150, California Division of
germination among Arctostaphylos stimulated by Mines, Sacramento.
heat and especially heat and charred wood extracts Emery, D. 1988. Seed Propagafion of Native Cafifornia
together, 4) more strongly enforced dormancy Plants. SantaBarbara Botanic Garden, SantaBarbara,
among Ceanothus spp. and 5) greater importance California.
of fire-suppressed germination. Conversely, germi- Griffin, J. R. 1978. Maritime chaparral and endemic
nation of the Adenostoma seed bank appears con- shrubs ofthe Monterey Region, California. Madrono
25:65-81.
sistent with what occurs elsewhere. Further study
of soil seed banks will be required to determine Hickman, J. C. (ed.). 1993. The Jepson manual: higher
plants of California. University of California Press,
whether there has in fact been divergence in the Berkeley.
germination ecology of maritime chaparral. In par- Holl, K. D., H. N. Steele, M. H. Fusari and L. R. Fox.
ticular, it would be illuminating to directly compare Seed banks of maritime chaparral and abandoned
seed banks of species that occur in both inland and roads: potential for vegetation recovery. Journal of
maritime chaparral. the Torrey Botanical Society 127:207-220.
Kaminsky, R. 1981. The microbial origin of the allelo-
Acknowledgments pathic potential ofAdenostomafasciculatum H. & A.
Ecological Monographs 51:365-382.
This research was supported by NSF Grant No. SES- Keeley, J. E. 1975. Longevity of nonsprouting Ceano-
8721494. James and Lauds Rose provided outdoor grow- thus. American Midland Naturalist 93:505-507.
ing space at their nursery. Frank Davis, Claudia Tyler, . 1977. Seed production, seed populations in soil,
Mark Borchert, Bruce Mahall, Carla D'Antonio, Bob and seedling producfion after fire for two congeneric
2000] ODION: GERMINATION OF MARITIME CHAPARRAL SEED BANK 203
pairs ofsprouting and nonsprouting chaparral shrubs. Santa Barbara County, California. Report prepared
Ecology 58:820-829. forSanta BarbaraCounty,ResourceManagementDe-
1984. Factors affecting germination ofchaparral partment.
.
seeds. Bulletin of the Southern California Academy Parker, V. T. 1987. Effects of wet-season management
of Sciences 83:113-120. burns on chaparral vegetation: implications for rare
. 1987. Role offire in seed germination ofwoody species. Pp. 233-237. in T. S. Elias and J. Nelson
taxa in California chaparral. Ecology 68:434-444. (eds.). Conservation and management ofrare and en-
. 1991. Seed germination and life history syn- dangered plants. California Native Plant Society, Sac-
dromes in the Californiachaparral. The Botanical Re- ramento.
view 57:81-116. , and V. R. Kelly. 1989. Seed banks in California
1992. Demographic structure ofCalifornia chap- chaparral and other Mediterranean climate shrub-
.
arral in the long-term absence offire. Journal ofVeg- lands. Pp. 231-255. in M. A. Leek, V. T Parker and
etation Science 3:79-90. R. L. Simpson (eds.). Ecology of soil seed banks.
. 1993. Utility of growth rings in the age deter- Academic Press Inc., San Diego, CA.
mination ofchaparral shrubs. Madroiio 40:1-14. Quick, C. R. 1935. Notes on the germination of Ceano-
, AND M. NiTZBERG. 1984. The role of charred thus seeds. Madrofio 3:135-140.
wood in the germination of the chaparral herbs E)n- , AND A. S. Quick. 1961. Germination of Ceano-
menanthe penduliflora (Hydrophyllaceae) and Erio- thus seeds. Madrofio 16:23-30.
phvilum confertiflorum {Asteraceae). Madroiio, 208- Sampson, A. W. 1944. Plant succession on burned chap-
218. arral lands in northern California. University ofCal-
, B. A. Morton, A. Pedrosa and P. Trotter. ifornia Agricultural Experiment Station Bulletin 685.
1985. The role ofallelopathy, heat and charred wood Shmida, a. and R. H. Whittaker. 1981. Pattern and bi-
in the germination of chaparral herbs and suffrutes- ological microsite effects in two shrub communities.
cents. Journal of Ecology 73:445-458. Southern California. Ecology 62:234-251.
, and C. J. Fotheringham. 1997. Trace gas emis- Stone, E. C. and G. Juhren. 1951. The effects offire on
sions and smoke-induced seed germination. Science the germination of the seed of Rhus ovata Wats.
276:1248-1250. American Journal of Botany 38:368-372.
, and . 1998. Smoke-induced seed germi- , AND . 1953. Fire stimulated germination.
nation in California chaparral. Ecology 79:2320- California Agriculture 7:13-14.
2336. Sweeney, J. R. 1956. Responsesofvegetation to fire. Uni-
Keeley, S. C, J. E. Keeley, S. M. Hutchinson and A. versity of California Publications in Botany 28:143-
W. Johnson. 1981. Postfire succession of the herba- 250.
ceous flora in southern California chaparral. Ecology Swank, S. E. and W. C. Oechel. 1991. Interactions
62:1608-1621. among the effects of herbivory, competition, and re-
, and M. Pizzorno. 1986. Charred wood stimulat- source limitation on chaparral herbs. Ecology 72:
104-115.
edgermination oftwo-firefollowingherbsoftheCal-
iforniachaparral andthe role ofhemicellulose. Amer- Thanos, C. a. AND P. W. RuNDEL. 1995. Fire-followers in
ican Journal of Botany 73:1289-1297. chaparral: nitrogenous compounds trigger seed ger-
Keith, D. A. 1987. Combined effects of heat shock, mination. Journal of Ecology 83:207-216.
smoke and darkness on germination ofEpacrisstuar- Tyler, C. M. 1995. Factors contributing to postfire seed-
tii Stapf., an endangered fire-prone Australian shrub. ling establishment in chaparral: direct and indirectef-
Oecologia 112:340-344. fects of fire. Journal of Ecology 83:1009-1020.
Mcpherson, J. K. and C. H. Muller. 1969. Allelopathic stud,iDe.s Co.f OMdoirorno aMnadnzDa.niMteaad{eA.rct1o9s9t8a.phEycloolsogmiocra-l
effects ofAdenostomcifasciciilatum, 'chamise\ in the
Californiachaparral. EcologicalMonographs39:177- roensis), seed ecology and reproductive biology. Re-
port prepared for California Department of Fish and
198. Game, Species Conservation and Recovery Program.
Moreenfofe,ctsJ.oMn.gearnmdinWa.tioCn. oOfecshherulb.s 1a9n9d1.heFribrseiinntseonustiht-y tors, affectin,g regeneraAtNiDonMo.fMA.orMrooriMtazn.za2n0i0t0a.(FAarcc--
MuLLeErRn,CaCl.ifHo.r,niRa.cBh.apHararnala.waEclotloagnyd7J2.:K1.99M3c-2p0h0e4r.son. tsopsotnaspehytloosprmeosrcrroiebnesdibsu)r:nirnegp.roRdeupctoirvtepbrieoplaorgeydafnordtrhee-
1968. Allelopathic control of herb growth in the fire California Department of Fish and Game, Species
cycle of California chaparral. Bulletin of the Torrey Conservation and Recovery Program.
Botanical Club 95:225-231. Wells, P. V. 1969. The relation between mode of repro-
Odion, D. C. 1995. Effects of variation in soil heating duction and extent ofspeciation in the woody genera
during fire on patterns ofplant establishment and re- ofthe California chaparral. Evolution 23:264-267.
growth in maritime chaparral. Dissertation, Univer- Wicklow, D. T. 1977. Germination response in Emmen-
sity of California, Santa Barbara. anthe penduliflora (Hydrophyllaceae). Ecology 58:
, AND F. W. Davis. 2000. Fire, soil heating, and the 201-205.
formation ofvegetation patterns in chaparral. Ecolog- Whelan, R. J. 1995. The ecology offire. Cambridge Uni-
ical Monographs 70:149-169. versity Press, Cambridge, Great Britain.
, D. E. Hickson, and C. M. D'Antonio. 1992. Wright, E. 1931. The effect ofhigh temperature on seed
Central coast maritime chaparral on Vandenberg air germination. Journal of Forestry 29:679-687.
force base: an inventory and analysis ofmanagement Zammit, C. a. and R H. Zedler. 1988. The influence of
needs for a threatened vegetation association. Report dominant shrubs, fire and time since fire on soil seed
prepared for The Nature Conservancy. banks in mixed chaparral. Vegetatio 75:175-187.
, J. Storrer and V. Semonsen. 1993. Biological , and p. H. Zedler. 1994. Organisation ofthe soil
resources assessment. Burton Mesa Project Area, seed bank in mixed chaparral. Vegetatio 11 1:1-16.