ebook img

Stage and Size Structure of Three Species of Oaks In Central Coastal California PDF

2014·10.6 MB·English
Save to my drive
Quick download
Download
Most books are stored in the elastic cloud where traffic is expensive. For this reason, we have a limit on daily download.

Preview Stage and Size Structure of Three Species of Oaks In Central Coastal California

Madrono, Vol. 61, No. 1, pp. 1-8, 2014 STAGE AND SIZE STRUCTURE OF THREE SPECIES OF OAKS IN CENTRAL COASTAL CALIFORNIA Ian S. Pearse Cornell Lab of Ornithology, 159 Sapsucker Woods Road, Ithaca, NY 14850 and Illinois Natural History Survey, 1816 S Oak St, Champaign, IL 61820 ianspearse@gmaiLcom Sophie Griswold and Desirree Pizarro Cornell Lab of Ornithology, 159 Sapsucker Woods Road, Ithaca, NY 14850 Walter D. Koenig Cornell Lab of Ornithology, 159 Sapsucker Woods Road, Ithaca, NY 14850 and NY Department of Neurobiology and Behavior, Cornell University, Ithaca, 14853 Abstract OaksarefoundationalspeciesinmuchofCalifornia, andmanyoakpopulationsinthestatemaybe injeopardydue to a lack ofrecruitment ofyoungtrees. Despiteconsiderable interestin this problem, therehavebeenfewcomprehensivesurveysofallstagesofoakdevelopment. Wesurveyedallstagesof threeoaks: Quercuslobata, Q. douglasii, and Q. agrifoUainaforestplotwithmixedland-useincentral coastal California. We found abundant seedlings ofall oak species, but an apparent paucity of Q. lobata and Q. douglasiisaplings. First year seedlings ofall species were less abundant in parts ofthe study site with cattle grazing, but later-stage seedlings of Q. lobata and Q. douglasii were equally abundant across land-use types. Quercus agrifolia seedlings were associated with non-grazed areas; Quercus agrifolia late-stage seedlings in the grazed area were smaller and less abundant than in non- grazedareas. Quercusagrifoliaseedlingsofallstagestendedtobeclusteredaroundconspecificmature trees. Quercus lobata late-stage seedlings, and to a lesser degree those of Q. douglasii, were often distant from any potential parent tree. These data indicate that young stages ofthe three species of oak have different spatial distributions and occur in different abundances at two sites with different grazing regimes. They are also consistent with a relative lack of regeneration in Q. lobata and Q. douglasii, although it remains to be determined that the small number of saplings of these species observed is insufficient to replace mortality of mature trees. This survey provides a baseline from which future resampling can assess the long-term demographic success of three Californian oak species. Key Words: regeneration, demography, Quercus. Californian oaks are a dominant component Callaway and Davis 1998; Davis et al. 2011), and of at least a third of the forest and savanna increased drought stress (Mahall et al. 2009). ecosystems of the state (Pavlik et al. 1991). Each of these factors has undergone widespread However, there has been recent concern for oak anthropogenic change in the last two centuries, populations, as there is a perceived lack of thus potentially explaining the apparently poor regeneration of oak forests and a lack of young recruitment currently suffered by oak popula- tree stages (Griffin 1976; Tyler et al. 2006). tions (reviewed by Tyler et al. 2006). However, Studies have attributed the lack of oak regener- the relative importance of each of these factors ation to various causal factors (Griffin 1971; may be site specific. For example, while grazing Borchert et al. 1989; Callaway and Davis 1998; may decrease seedling abundance at some sites Gordon and Rice 2000), while others have (Reiner and Craig 2011), even sites without questioned the severity of regeneration problem livestock grazing may experience a lack of altogether (Tyler et al. 2006). seedling recruitment (Griffin 1971). The lack of The apparent lack of recruitment of two a single causal factor in altering oak populations common California oaks, Quercus lobata Nee calls for a more integrative approach in assessing and Q. douglasii Hook. & Arn, (Fagaceae), have oak population dynamics. prompted numerous causal explanations for A recent conceptual review (Tyler et al. 2006) seedling mortality and lack of regeneration in and meta-analysis (Zavaleta et al. 2007) of these species, including cattle grazing (Hall et al. California oak recruitment both suggested that 1992; Tyler et al. 2008; Reiner and Craig 2011), two Californian oaks, valley oak {Q. lobata) and competition from non-native grasses (Griffin blue oak {Q. douglasii), suffer from declining 1971; Gordon and Rice 2000), altered herbivore populations. Both studies concluded that site- populations (Griffin 1971; Borchert et al. 1989; specific causes may affect recruitment of these MADRONO 2 [VoL 61 species. Moreover, they indicated that detailed Methods demographic models are needed to address the question of whether low observed recruitment Study Area and Natural History levels are actually insufficient to replace stands of long-lived adults. Likewise, a detailed demo- The survey was conducted on Haystack Hill graphic analysis of oak populations could ask andsurroundingareas(36.385114N, —121.561906 specifically which life stages most limit popula- W) on Hastings Natural History Reservation tion growth (Tyler et al. 2006; Zavaleta et al. (HNHR) and the adjacent Oak Ridge Ranch 2007; Davis et al. 2011). Meanwhile, studies have (ORR) (Fig. 1). The study area is typical ofcoast suggested that a third species ofCalifornian oak, range oak habitat, is the site of several studies of Q. agrifolia Nee, may not be undergoing similar oak populations (White 1966; Griffin 1971, 1976; cloonwgenreergse,nereavteinonthroautgehs iatsmiatsy tbweoafsfyemctpeadtrbiyc K70oeknmigofa1n/d3 oKfntohpesst2u0d0y7)s,iteas)ndtoitshecesnittresalus(ewditihnian some of the same recruitment challenges (Call- recent meta-analysis of Californian oak popula- away and Davis 1998; Mahall et al. 2009). tions (Zavaleta et al. 2007). The study area has a While studies have successfully documented MediterraneanclimatetypicalofCalifornia’scoast multiple site specific factors likely affecting range. Intheyearofthesurvey,theaveragerainfall rpeolpautliavetiloancsk ooffCtahleifexoprlniicaint doeakmso,grtahpehreicismostdilell-a w30asye4a7r5avmemr,agewhfoirchthweassitesl(i5g4h0tlymmless- tHhaasntintghse Weather Station http://www.hastingsreserve.org/ ing of oak populations needed to quantify the weather/Weather.html). Thestudyareawasrough- sritsakgetsooofakoapkospualraetioofnsthaendhitgoheasstsescsonwsheirvcahtiliofne lydelimitedbythecontourofthehill. Weexcluded a large patch ofdense chaparral that would have value. The only current study that explicitly been within the study, as the area was too densely reconstructs Q. lobata demography uses an vegetated to use the same survey techniques as in experimental approach, which is very valuable other areas and contained few oak trees, with the in determining causal factors that limit Q. lobata exception of a small number of Q. agrifolia success (Davis et al. 2011). Observational studies individuals at the bottom of dry washes that that thoroughly survey oak populations that are descended the hillside through chaparral. both geographically referenced and include all life The study area was divided into three habitat stages of the oaks complement experimental types based on tree cover and groundcover type: studies. Such surveys will be valuable because forest, savanna, and grassland (Griffin 1990). they can be easily resampled in order to directly ‘Forest’ had a complete canopy cover. Forest assess vital rates and stage transitions. At the understory was dominated by poison oak {Tox- same time, comprehensive site surveys observe icodendron diversilobum [Torr. & A. Gray] oak populations across all microhabitats at a site, Greene) and had little to no grass cover. One which is difficult in manipulative studies. forest region had a substantial stand ofmadrone Here we present a baseline survey of the oaks {Arbutus menziesii Pursh), whereas oaks domi- on a portion ofHastings Natural History Reserve nated all others. ‘Savanna’ had a partial canopy (HNHR) and the adjacent Oak Ridge Ranch coverand an understory ofnative and non-native (ORR) property in Monterey County of central grasses including Avena L. sp., Bromus hordea- coastal California. We surveyed the size, stage, ceus L., R diandrus Roth, Hordeum L. sp., Stipa and geographic location of 19,755 individuals of pulchra Hitchc., and Aira caryophyllea L., as well the three dominant oak species {Q. lobata, Q. as forbs such as Madiagracilis (Sm.) D. D. Keck, douglasii, and Q. agrifolia) in a 52.6 hectare area Asclepias eriocarpa Benth., Plagiobothrys Fisch. encompassing forest, grassland, and savannah & C.A. Mey. sp., and Amsinckia menziesii habitats. This survey also encompassed areas (Lehm.) A. Nelson & J.F. Macbr. ‘Grasslands’ with strikingly different recent land use histo- had sparse to no tree cover and were dominated HNHR ries, as has been protected from cattle by the same mix of grasses and forbs. Small ORR grazing since 1937, but continues to remnants of perennial grasslands (sensu Griffin experience significant regular cattle grazing. 1990) dominated by Stipa pulchra were included We use this survey primarily to establish a in this category. There wasconsiderable variation baseline for parameterizing future demographic in both slope and aspect of habitat types within models of oak populations, and we make this the study area, but both study sites (HNHR and baseline data available to the general scientific ORR) had both steep and relatively flat forests, community. We analyzed the size structure and grasslands, and savannahs. One difference be- spatial distribution of early and late seedling tween the two sites was the presence of a small HNHR stages of these oaks in order to ask to what creek bed at the site that had a high degree habitat type and land use affect the size abundance of Q. agrifoUa seedlings. of late-stage seedlings and the proximity of The study site was divided approximately HNHR seedlings to a potential parent. down the middle by a fence separating 2014] PEARSE ET AL.: SURVEY OF CENTRAL CALIFORNIAN OAK STAGES 3 Fig. 1. A map of the survey area (highlighted in light dashes and outlined with a dotted line). The survey is divided roughly in halfby a property line (central E~W line), where the northern property (Oak Ridge Ranch) is grazed andthesouthernproperty (HastingsNatural History Reserve (HNHR) isungrazed. In May-June 2012, we surveyed and recorded thelocation ofallvalley oak (Q. lobata, white dots), blue oak (Q. douglasii, greydots), and coast live oak (Q. agrifolia, black dots) first year seedlings, late-stage seedlings, saplings, and trees within the demarcated study area. from ORR. The land use ofthe two areas can be transects across the study area and used a GPS to summarized as follows. Prior to 1930, both sites recordthelocalityofeachoakwithintheareatoan had a history of mixed land use including accuracy of5 m. For each individual, we recorded ranching, timber harvest, and small-scale agricul- one offourgrowthstages. Seedlingswerefirstyear HNHR ture (Griffin 1990). was donated to the germinants from acorns. All had unbranched UniversityofCaliforniain 1937 atwhichpointall stems, and most were still connected to a partially grazing and agriculture was stopped. At ORR, buried acorn. Late-stage seedlings were over one moderate year-round cattle grazing continues to year old, less than 1.5 meters in height, had present day, including during the active survey branched ormultiple stems, and lacked a connect- period, although no tillage or significant timber ed acorn. Saplings were greater than 1.5 meters in harvest has been recorded within the study area height, less than 10 cm DBH, and had not yet within recent history. While oak plantings and beguntoproducefruitorflowers. Adulttreeswere restoration efforts have been made in other over 10cmDBH.Weconsideredtheseclassestobe portions ofHNHR, none ofthese activities have biologically useful, as demographic parameters occurred within the study area. includinggrowthrate, survivorship, andreproduc- The year preceding our survey had a moderate tion likely vary between stages (e.g., Griffin 1971; togoodacorncropforallthreeoakspecies. Based Koenig and Knops 2007). For all late-stage on the Koenigvisual countmethod (Koenigetal. seedlings, we also recorded height and canopy 1994a), the 2011 Q. lobata acorn crop was 15% width and calculated seedling volume as height X greater than the 30-year mean Q. lobata crop widtffi. For all first-year seedlings, late-stage (Pearse et al. 2014). Likewise, the 2011 Q. seedlings, and saplings, we also calculated Euclid- douglasii acorn crop was 63% greater than its ean distance to the nearest conspecific adult tree. 30-year mean, and the 2011 Q. agrifolia acorn Wetaggedeachadulttreeandrecordedwhichfirst crop was 29% greater than its 30-year mean. year seedlings, late-stage seedlings, and saplings m were within 20 ofits trunk. Survey Methods Statistical Analysis We surveyed allvalley oak, blue oak, andcoast live oak individuals within the study area in May While the two landscape level factors - habitat through early July 2012. We marked 10 m wide and property, the latter corresponding to differ- MADRONO 4 [VoL 61 Table 1. The Density of Three Oak (Quercus) Species (Trees/Hectare) in Three Habitat Types Spanningthe Ungrazed Hastings NaturalHistory Preserve(HNHR)andGrazed OakRidgeRanch (ORR). All trees (>10 cm DBH), saplings (>1.5 m tall), late-stage seedlings (multiple stems, no attached acorn), and first-year seedlings (single stem, often acorn still attached) within the total 52.57 ha survey areawere recorded resulting in 19,755 records ofoak individuals. Forest Savannah Grassland HNHR ORR HNHR ORR HNHR ORR Area (ha) 8.21 8.89 8.23 7.73 10.83 8.65 Quercus lobata 1st year seedling 161.75 7.2 49.09 5.3 16.34 0.55 Late stage seedling 47.14 17.55 24.79 54.59 15.51 18.47 Sapling 0.24 0 0 0 0.09 0.09 Tree 17.54 10.01 13.49 6.47 2.4 0.37 Quercus douglasii 1st year seedling 18.39 176.04 84.93 34.67 3.23 0.09 Late stage seedling 29.6 194.94 94.9 60.28 5.26 2.4 Sapling 0.49 0 0 0 0.09 0 Tree 11.33 69.74 22.72 6.21 0.09 0.83 Quercus agrifolia 1st year seedling 455.18 23.17 57.84 1.29 61.68 0.28 Late stage seedling 209.38 76.83 24.18 8.8 6.74 3.42 Sapling 15.23 0.45 0 0 0.55 0 Tree 55.05 27.67 6.44 0.91 0.92 0 ent land use - are spatially non-independent, we with savannah, although the distribution of used these data to build a case study of oak overstory trees only partially reflected this trend. populations at a single, large site. Specifically, our All growth stages of Q. douglasii were uncom- goal was to explore variation in 1) the size oflate- mon ingrasslands at all sites (Table 1). Therewas stage seedling oaks, as this size class may escape a greater abundance ofearly growth stages of Q. ORR HNHR grazing at large sizes (either by being tall or very douglasii at than at (4059 versus wide), 2) the distance of seedlings and late-stage 1966), reflecting the distribution ofadult trees at seedlingoaksfromapotential (conspecific)parent the two sites (677 versus 281). There was a full- tree, and 3)thecompositionofoakseedlingsinthe canopied Q. douglasii forest on ORR, and all understory of individual adult trees. All statistics growth stages of Q. douglasii at ORR were more were calculated in R using package car (R Core closely associated with forest than at HNHR, Development Team 2008; Fox and Weisberg where Q. douglasii was more abundant in open 2011). Our georeferenced dataset is made avail- savannah habitat. Quercus agrifolia was associ- able at the Dryad data depository http://dx.doi. ated with full-canopied forest habitats at both HNHR org/10.5061/dryad.467g5 and on the authors’ sites and was particularly abundant in websites http://www.nbb.cornell.edu/wkoenig; forests. Quercusagrifoliawas also more abundant HNHR http://ianpearse.wordpress.com. in grasslands and savannas than the same habitat types at ORR. There were strikingly Results fewer Q. agOriRfoRlia first yeaHrNHseRedlings in all habitats on than on (219 versus Habitat and Land-use Associations 4881), which would be consistent with an impact of grazing on Q. agrifolia seedling abundance. We surveyed 52.6 ha and gathered data on a For Q. agrifolia, this trend persisted into late- total 19,755 individuals of the three oak species stage seedlings, where there were fewer Q. acrossboth properties (HNHR andORR) and all aHgrNifHolRia late-stage seedlings at ORR than at habitats (Fig. 1, Table 1). The surveyed area of (788 versus 1991). each habitat type at each property was similar, There were few saplings ofboth Q. lobata and andwereportthedensitiesofoaksateachstagein Q. douglasiiinallhabitatsatbothproperties(four each habitattype on the two properties (Table 1). Q. lobata saplings and five Q. douglasiiseedlings), Quercus lobata seedlings were more abundant consistent with perceived recruitment problems at HNHR, the ungrazed site, than at ORR, the with these species. In contrast, Q. agrifolia grazed site - 1909 seedlings versus 111 (Table 1). saplings were reasonably common (n = 135), However, this difference did not carry into the almost all of which wHerNeHaRssociated with forest late-stage seedling individuals, and there were sites on the ungrazed property (Table 1). slightly fewer late-stage seedling Q. lobata at HNHR than at ORR (759 versus 778). At Size of Late Stage Seedlings HNHR both Q. lobata seedlings and late-stage seedling stages tended to be located in forested We measured the volume of late-stage seed- habitats, whereas at ORR, they were associated lings (excluding first-year seedlings) (Fig, 2). 2014] PEARSE ET AL.: SURVEY OF CENTRAL CALIFORNIAN OAK STAGES 5 stage seedlings in grasslands than in otherhabitat types (Fig. 2). Unlike Q. lobata, they were roughly the same size across land-use types in all habitats. Quercus agrifolia late-stage seedlings showed the strongest size association with land use type of the three species. In both grasslands and HNHR forests, Q. agrifolia late-stage seedlings at were 300-400% larger than those in the same ORR habitats on (Fig. 2). Quercus agrifolia late- C0/55 forest savannah grassland stage seedlings in savannah habitat were smaller g than in other habitats (Fig. 2). The variation in size of seedlings was greatest in grassland 0 0 habitats. (/) 0 Spatial Distribution of Seedlings 05 S to We measured the distance from each first year BI seedling and late stage seedling to its nearest mature conspecific tree (i.e., potential parent) o (Fig. 3). Quercus lobata first-year seedlings aver- aged 12.4 meters from the nearest potential parent. Quercus lobata late stage seedlings were on average further from the nearest potential forest savannah grassland parentthan seedlings ofthe otherspecies, andthis ORR distance was even greater on than on HNHR m m (15.6 and 14.3 respectively, Fig. 3). Quercus douglasii seedlings were generally closer to a potential parent than Q. lobata seedlings, Q. agrifolia likely reflecting the density ofstands ofthese tree species (Fig. 3). Both Q. douglasii first-year and m late-stage seedlings averaged eight from the nearest potential parent on HNHR (Fig. 3). On ORR, first-year Q. douglasii seedlings averaged m 4.7 from the nearest potential parent and late- stage seedlings averaged seven m. Quercus agrifo- m lia seedlings averaged 7.5 from the nearest potential parent, a value that did not vary by forest savannah grassland either seedlings class (Fi 3855 =2.0, P = 0.16) or property (Fi^gss =0.002, P = 0.96, Fig. 3). Fig. 2. The volume of late-stage seedlings of Q. We also measured the composition of the tloybpaetsa,onQ.udnogurglaazseidi,(aHnNdHQR.,agdraifrokliagrieny)tharnede hgarbaizteadt seedling oak community underneath mature trees (ORR, light grey) properties. Our designation of“late- of each of the three oak species and found that stage seedling” included oaks that were less than 1.5 m the understories of both Q. douglasii and Q. in height, had multiple stems, and no attached acorn. agrifolia were dominated by their conspecific This size ofthis stage may be particularly important to seedlings (Fig. 4). In contrast, the understories of oak regeneration, as late-stage seedlings persist over Q. lobata trees had a roughly even distribution of many years and are susceptible to grazing, especially seedlings from all three oak species. Similarly, Q. when small. Bars are means ± SE. lobata seedlings were the least abundant compo- nent ofthe understories of either Q. douglasii or Quercus lobata late-stage seedlings tended to be Q. agrifolia. larger in more open habitats such as grasslands irrespective ofwhether the site was grazed or not Discussion (Fig. 2). Quercus lobata seedlings in savannahs were four times larger on ORR than on HNFIR The size and age distributions of the three (Fig. 2) and similar in size in other habitat types, species of oaks surveyed in this study varied a finding that is not consistent with cattle grazing between properties with different land-use histo- being a factor limiting seedling growth of this ry, but these differences were fairly subtle. There species. were fewer valley oak {Q. lobata) first-year Quercus douglasii late-stage seedlings showed seedlings per adult tree on the grazed site similar size trends to Q. lobata, with larger late- (ORR) than on the ungrazed site (HNHR). MADRONO 6 [Vol. 61 Fig. 3. ThedistributionofthedistanceofQ. lobata, Q. douglasii, and Q. agrifoliafirstyearseedlings(dashedline) and late-stage seedlings (solid line) to their nearest conspecific (i.e., potential parent) tree at the ungrazed site (HNHR) and grazed site (Oak Ridge Ranch). Valley oak late-stage seedlings and to a lesser degree first year seedlings are located at a greater distance from a potential parent than the other two species. However, the abundance of late-stage seedlings Interestingly, our survey found fewer (Table 1) was similar at both sites (Table 1). Moreover, Q. and smaller (Fig. 2) early-stage Q. agrifolia lobata late-stage seedlings were on average larger individuals on the grazed property (ORR) than at the grazed site, further suggesting that cattle on the ungrazed property (HNHR). Unlike in Q. grazing may affect the survival of first-year douglasii, Q. agrifolia later-stage seedlings were seedlings, but may have little or no effect on the both smaller and less abundant (with reference to ORR HNHR long-term survival ofestablished seedlings. Quer- mature trees) on than on (Fig. 2, cus douglasii populations had generally the same Table 1). Moreover, there were far more Q. HNHR associations as Q. lobata with the two major land agrifoUa saplings per mature tree on use patterns (Table 1, Fig. 2), except that there than on ORR (Table 1). As noHn-NadHuRlt stages was a large number offirst-year and late-stage Q. appear to be more represented at than at douglasii seedlings in a dense forest patch on ORR, this suggests that Q. agrifolia population ORR (Table 1). The larger size of late-stage growth on HNHR may be higher than on the seedlings at the grazed site suggests that cattle grazed adjacent property. Quercus agrifoUa pop- grazing does not limit (and may even promote) ulations are thought to be stable or even the growth ofthese seedlings at our study area. increasing in many regions with high forest or The land-use differences in this survey are shrub cover, but are unable to colonize open consistent with effects of cattle grazing on Q. habitat, presumably because shrubs provide a douglasii and Q. lobata observed in other studies. refuge from wildlife or livestock grazing (Call- For example, experimental manipulation of away and Davis 1998). In the current survey, Q. grazing did not affect the survivorship of late- agrifoUa was strongly associated with habitats stage (2-year old) seedlings in a Sierra Nevada with greater shrub and tree cover, and livestock foothills population of Q. douglasii. (Hall et al. grazing may have imposed a sufficient herbivore 1992), and grazing decreased the total abundance pressure to partially negate the protective effect of naturally recruiting Q. douglasii seedlings ofhighvegetation coverfor Q. agrifolia seedlings. (likely including a large portion of first-year The seedlings of each of the three oak species seedlings) in only one out of four years (Reiner had unique spatial distributions with reference to and Craig 2011). Livestock grazing may still be a their potential parental trees. The distribution significant mortality factor for Q. lobata and Q. of first-year seedlings likely reflects patterns of douglasiv, however, our results suggest that other acorn fall and immediate dispersal. As acorn mortality factors are potentially more important production is pulsed (i.e., varies dramatically even in ungrazed sites. from yearto year) and synchronous between trees 2014] PEARSE ET AL.: SURVEY OF CENTRAL CALIFORNIAN OAK STAGES 7 Q. douglasii some sites, resulted in an underrepresentation of Q. lobata seedlings in the understory ofQ. lobata mature trees (Fig. 4). Quercus douglasii seedlings were more abundant under their parent tree (Fig. 4), although late-stage seedlings tended to be located further from a potential parent than first-year seedlings (Fig. 3). Both first year and late-stage Q. agrifolia seedlings were strongly clustered around potential parental trees (Fig. 3), and Q. agrifolia seedlings were highly represented in the understories of Q. agrifolia trees (Fig. 4). These results suggest a hierarchy in the ability of oak seedlings to colonize open habitat. Quercus lobata was often found far from its Q. agrifolia parental species, consistent with previous studies showing that Q. lobata seedlings better tolerate dry habitats (Mahall et al. 2009) and deer browsing (Tyler et al. 2008) better than Q. agrifolia. Quercus douglasii has been shown to tolerate dry, exposed habitats (Griffin 1971), but it may be negatively affected by direct competi- tion from non-native grasses in open areas (Gordon and Rice 2000) and is associated with shrub cover (Callaway 1992). Quercus agrifolia has been shown to be strongly associated with shrubby or forested habitat, which provides protection from herbivores (Callaway and Davis 1998). The factors that limit oak populations are thought to be somewhat site-specific. This has led to both studies that attempt to find those factors that are important across many site (Zavaleta et al. 2007) and detailed studies that attempt to assess the causal factors limiting oak seedling success at individual sites (Davis et al. 2011). Our study will be useful in both of these goals. By establishing a baseline survey at a site, where the historyoffactors thatcouldpotentiallyaffect oak populations (e.g., grazing, habitat affiliation, wildlife density) is well documented, our study yFeiag.r 4a.ndTlhaete-sspteacgieessiedeednltiintgys)ofwialtlhiinmm2a0tumreoofaaksva(flilresyt can easily be combined with other surveys at oak(Q. lobata = Q. lo),blueoak{Q. douglasii= Q. do), different geographic locations in order to assess and coast live oak {Q. agrifolia = Q. ag). The the generality of these factors. Likewise, by understories of both blue oak and live oak were having a population of oaks where all stages are dominated by conspecifics (likely acorn fall from the surveyed and georeferenced, this surveyopens the focal tree), however the understory ofvalley oak had a possibility of resampling to determine the key roughly even representation ofall three species. demographic limitations at this representative site. (Koenig et al. 1994b), the distribution of first- Explicit demographic models of Californian year seedlings may vary substantially from year oak populations are important to understand the to year. On the other hand, the late-stage sustainability of the state’s oak populations seedlings of oaks in this habitat may be very (Davis 2011). This study provides a thorough long-lived (Koenig and Knops 2007), so their baseline estimate ofpopulations of Q. lobata, Q. distribution is unlikely affected by temporal douglasii, and Q. agrifolia that can be used by variation in acorn crop. Quercus lobata first year future resampling and modeling efforts to esti- seedlings were clustered around potential paren- mate demographic transitions between life-histo- tal trees, likely reflecting patterns of acorn fall. ry stages and thus the long-term dynamics ofoak Late-stage Q. lobata seedlings, however, were on populations. For example, at the same site, a average further from a potential parent, especial- marked population of blue oak late-stage seed- ly on the grazed property (Fig. 3), This, in lings has persisted with little growth, little combination with few seedlings altogether at mortality, and no transitions to saplings within MADRONO 8 [Vol. 61 the past 60 years (Koenig and Knops 2007). blue oak seedling damage and survival. Journal of Moreover, there arecurrentlyno robust estimates Range Management 45:503-506. of age-biased mortality for adult valley oaks, Koenig, W. D. andJ. M. H. Knops. 2007. Long-term which is a critical parameter in estimating the growth and persistence of blue oak {Quercus necessary rate ofreplacement by seedlings (Tyler douglasii) seedlings in a California oak savanna. aestanpdalhinsg2es0e,0d6lr)i.ensgBasm,yplmaiasnrgkwieonlfgltahaislsarstgtheuednyufmembwaeyroboosbfestrervreevedes uRMs.aid,nLrg.ovniMosuua5ml,4mW:s2.eu6.r9Jv.-e2yC17sa94.9r.4mCaae.nna,EdsiMtai.nmaTt.JioSnutgrnaanalbcoaorcfnk,Fcoarroenpsdst enough stage-transitions to accurately parame- Research (Revue Canadienne De Recherche For- terize demographic models ofoak populations at estiere) 24:2105-2112. a large, natural site. , R. L. Mumme, W. J. Carmen, and M. T. Stanback. 1994b. Acorn production by oaks in Acknowledgments central coastal California: variation within and WethankVinceVoegeliandthemanagementofOak among years. Ecology 75:99-109. sRuipdpgoerRteadncbhyfaonraNccSeFssRtoEtUhegsrtaundtyssDitEe.BT-hi1s02s1t4u1d7ywaansd MahaMlalt,a.B.200E9.,. AC.coMm.paTryaltievre,sEt.udyS.ofCoolaek,{QaunerdcuCs., DEB-1212885 and NSF grant DEB 0816691 to WDK. Fagaceae) seedling physiology during summer drought in southern California. American Journal Literature Cited ofBotany 96:751-761. Pavlik, B. M., P. C. Muick, S. G. Johnson, and M. Borchert, M. L, F. W. Davis, J. Michaelsen, and Popper. 1991. Oaks ofCalifornia. Cachuma Press L. D. Oyler. 1989. Interactions of factors and California Oak Foundation, Los Olivos, CA. affecting seedling recruitment ofblue oak {Quercus Pearse, I. S., W. D. Koenig, and J. M. H. Knops. douglasii) in California. Ecology 70:389-404. 2014. Cues versus proximate drivers: testing the Callaway, R. M. 1992. Effect of shrubs on recruit- mechanism behind masting behavior. Oikos (in ment of Quercus douglasii and Quercus lobata in press). California. Ecology 73:2118-2128. R Development Core Team. 2008. R: A language and F. W. Davis. 1998. Recruitment of and environment for statistical computing. R Quercus agrifolia in central California: the impor- Foundation for Statistical Computing, Vienna, tance of shrub-dominated patches. Journal of Austria. Website http://www.R-project.org/ (ac- Vegetation Science 9:647-656. DaviCs,onFs.uWm.e,rC.cMo.ntTroyllero,faonadkB.deE.moMgarhaaplhly. 2i0n11.a Reinceerss,edRD.eca2n0d10)A.. Craig. 2011. Conservation Mediterranean-climate savanna. Ecosphere 2:108, easements in California blue oak woodlands: testing the assumption of livestock grazing as a doi:10.1890/ESl1-00187.1 Fox,ApJ.plainedd RS.egWreeisssiboenr,g.Se2c0o1n1d. AEdnitRionC.omSpaageniPounblit-o Tylecro,mpCa.tiMb.l,eFu.seW..NaDtauvriasl,AarnedasB.JoEu.rnMaalh3a1l:l40.8-2401038.. cations, Thousand Oaks, CA. The relative importance of factors affecting age- Gordon, D. R. and K. J. Rice. 2000. Competitive specific seedling survival of two co-occurring oak suppression of Quercus douglasii (Fagaceae) seed- species in southern California. Forest Ecology and ling emergence and growth. American Journal of Management 255:3063-3074. Botany 87:986-994. , B. Kuhn, and F. W. Davis. 2006. Demogra- Griffin, J. R. 1971. OakregenerationinupperCarmel phy and recruitment limitations of three oak Valley, California. Ecology 52:862-868. species in California. Quarterly Review ofBiology . 1976. Regenerationin Quercuslobatasavannas, 81:127-152. Santa Lucia Mountains, California. American White, K. L. 1966. Structure and composition of Midland Naturalist 95:422-435. foothill woodland in central coastal California. 1990. Flora of Hastings Reservation, Carmel Ecology 47:229-237. . Valley, California. Third Edition. Regents of the Zavaleta, E. S., K. B. Hulvey, and B. Fulfrost. University ofCalifornia, Berkeley, CA. 2007. Regional patterns ofrecruitment success and Hall, L. M., M. R. George, D. D. McCreary, and failure in two endemic California oaks. Diversity T. E. Adams. 1992. Effects of cattle grazing on and Distributions 13:735-745.

See more

The list of books you might like

Most books are stored in the elastic cloud where traffic is expensive. For this reason, we have a limit on daily download.