Advertisement

Plant Ecology

, Volume 220, Issue 2, pp 151–169 | Cite as

Drought in Southern California coastal sage scrub reduces herbaceous biomass of exotic species more than native species, but exotic growth recovers quickly when drought ends

  • Chandler E. Puritty
  • Ellen H. Esch
  • Sherlynette Pérez Castro
  • Elizabeth M. Ryan
  • David A. Lipson
  • Elsa E. ClelandEmail author
Article

Abstract

Semi-arid regions with Mediterranean-type climates harbor exceptional biodiversity, but are increasingly threatened by invading exotic annual species and climatic changes, including drought. In semi-arid ecosystems, antecedent conditions often influence plant growth, but the role of antecedent conditions for drought response and recovery of native versus exotic species remains largely unexplored. From 2013 to 2016, we imposed experimental rainfall treatments (average rainfall, moderate or severe drought) in plots under a native shrub canopy and in inter-spaces dominated by herbaceous vegetation, and quantified growth (peak biomass) and abundance (cover) of native and exotic herbaceous species. The following year, we quantified recovery from the drought treatments (2017). Exotic biomass was less resistant to drought (declined more than native biomass), but was more resilient (increased more than native biomass in the year following drought), especially in unshaded inter-spaces between shrubs. These responses were associated with life history; annual species responded more negatively to drought in the inter-spaces than perennial species. Current years' rainfall was a better predictor of biomass than prior rainfall, but antecedent factors were also important. After four years of rainfall treatments, exotic species had the highest growth recovery in the severe drought treatment, while growth of natives had the opposite response. In contrast, litter was positively associated with plant growth regardless of origin. This study demonstrates that when native and exotic species differ in life history, as they do in Mediterranean climate ecosystems, they may respond differently to antecedent factors, and hence differ in recovery from climate extremes such as drought.

Keywords

Antecedent conditions Ecological resilience Invasion Litter Seedbank 

Introduction

The California floristic province is a global biodiversity hotspot with high rates of endemism, especially in the Mediterranean climate region of the state (Cincotta et al. 2000). This region is also at risk for species’ loss due to invasion by exotic species (Seabloom et al. 2006; Underwood et al. 2009) and climate change (Loarie et al. 2009; Harrison et al. 2015). Significant climate change is predicted for this region in the coming decades (Cayan et al. 2008; Diffenbaugh and Giorgi 2012), including drought (Seager et al. 2007; Diffenbaugh et al. 2015) and increased interannual variability in rainfall (Berg and Hall 2015; Yoon et al. 2015). Local-scale studies often observe that high rainfall increases the growth or abundance of exotic annuals (Hobbs et al. 2007; Ashbacher and Cleland 2015). Meta-analyses have shown that invasive species often have higher leaf area and growth rates than non-invasive species (Van Kleunen et al. 2010), as well as traits associated with higher water demand (Cavaleri and Sack 2010). Accordingly, exotic invasive species often have higher performance under conditions of high resource availability (including water) than native species (Daehler 2003). Similarly, recent meta-analyses have found that exotic species tended to respond more negatively to drought than native species (Sorte et al. 2013; Liu et al. 2017). Together these studies suggest that drought should impact growth of native species less than exotic species.

However, regional models project that some exotic annual grasses in the semi-arid regions of the Western U.S. will expand their ranges with declining rainfall (Bradley 2009). One way that drought might promote invasion is by reducing native biomass, thus reducing competition for light, and soil resources, and benefiting fast-establishing invasive exotic species in subsequent years with higher rainfall (Jiménez et al. 2011; Diez et al. 2012). Drought can also promote fire (Dennison et al. 2014), and positive feedbacks by facilitating exotic grass invasion, fine fuel accumulation, and hence accelerated fire regimes (D'Antonio and Vitousek 1992; Balch et al. 2013). Given the impacts of invasions on ecosystem processes and biodiversity (Liao et al. 2008; Ehrenfeld 2010; Powell et al. 2011; Vilà et al. 2011), it is critical to better understand how invasion by exotic species will influence ecosystem responses to drought and interannual rainfall variation in this region.

Most of the studies evaluating variation among functional groups in their response to drought have focused on relationships with rainfall in the most recent growing season. However, antecedent environmental conditions can play a large role in predicting ecological processes, resulting in a time lag in correlations between environmental conditions and plant growth (Ogle et al. 2015). For instance, plant growth and production can be influenced by rainfall in prior years, in addition to the current year (Lauenroth and Sala 1992; Sala et al. 2012). Prior years’ rainfall is particularly important for understanding growth dynamics in semi-arid systems, including grasslands (Oesterheld et al. 2001) and shrublands (Anderson and Inouye 2001). In semi-arid systems, high rainfall can increase seed bank abundance (Holmgren et al. 2006; Dudney et al. 2017) and perennial meristems (Reichmann et al. 2013), contributing to an increase in plant growth in the following year. However, following a year of high production, standing litter can increase soil shading, which can have either positive or negative effects on seedling establishment, growth, and community-level production in subsequent years (Facelli and Pickett 1991; Wolkovich et al. 2010). Prior work in herbaceous-dominated ecosystems has shown that plant functional groups vary in response to antecedent conditions (Dudney et al. 2017), and resilience following drought (Hoover et al. 2014), but few studies have investigated how native versus exotic species vary in this respect. In notable exceptions, Concilio et al. (2016) found that introduced cool-season species responded more to antecedent rainfall than native species in a semi-arid grassland in Colorado, and Potts et al. (2006) found differing responses of ecosystem carbon flux to antecedent conditions depending on whether native or invasive species dominated. However, it remains unclear whether the results of these initial studies are transferrable to other regions and vegetation types.

Here, we evaluated the drought response and recovery of biomass and abundance (cover) of native and exotic herbaceous species, as well as the role of antecedent conditions in these responses. We conducted our research in Californian coastal sage scrub (CSS), a semi-arid ecosystem traditionally dominated by drought deciduous shrubs and diverse herbaceous species (Rundel 2007; Cleland et al. 2016). However, CSS is increasingly invaded, especially by exotic annual species (Minnich and Dezzani 1998). Soils in CSS can harbor a large seedbank of exotic annual species, enabling these exotic species to establish quickly following disturbance (Cox and Allen 2008). In Southern California, nitrogen deposition and accelerating fire frequency in CSS have been associated with type-conversion from shrubland to grassland dominated by exotic annual species (Talluto and Suding 2008). Hence, we evaluated how herbaceous species responded to drought under the shrub canopy, as well as in the inter-spaces between shrubs, the two contexts where herbaceous species are found in this system. As with other Mediterranean climate ecosystems (Funk et al. 2016), invading species are more likely to be annual than perennial in this system. Hence, it is important to note that the differential responses of native and exotic species in this system are likely tied to life history differences, and associated variation in growth and response to resource variation (Chapin 1980; Grime 1977; Tilman 1985).

From 2012 to 2016, California experienced the most severe drought in millennia (Griffin and Anchukaitis 2014; Robeson 2015). During these 4 years, we maintained a rainfall manipulation experiment (average rainfall, moderate or severe drought treatments) in adjacent plots either under a native shrub canopy or in inter-spaces dominated by herbaceous vegetation, and then ceased treatments and monitored biomass during the first high rainfall year following the drought (2017). We expected that (1) exotic species growth would decline more with drought than native species, because exotic species in this system are mostly annual and have a suite of traits including fast growth but low tolerance for moisture stress. We further expected that (2) biomass of both native and exotic species in the recovery year following drought would depend on antecedent factors such as rainfall in prior years, abundance of seeds in the seedbank, and standing litter from prior years' biomass. A synthesis of long-term observations in the United States found that annual species were more sensitive than perennial species to interannual rainfall variability (Cleland et al. 2013), thus we expected that biomass of annual exotic species would be more sensitive to these antecedent factors than biomass of perennial native species. In contrast, we expected that (3) the seedbank would be of greater importance for understanding the response of native biomass to drought compared to exotic biomass, where seeds of native species would be more abundant in the seedbank following drought than seeds of exotic species, as has been found elsewhere in California (LaForgia et al. 2018). Finally, we expected that (4) antecedent factors could play different roles in drought recovery of herbaceous biomass under the shrub canopy than in the shrub inter-spaces, given that shade and antecedent factors such as litter accumulation under the shrub canopy could potentially buffer plants from desiccation under drought conditions.

Methods

A rainfall manipulation experiment was conducted from 2012 to 2017 in the Santa Margarita Ecological Reserve (33°29′N, 117°09′W) located at the Riverside-San Diego county line, California, USA. The site has a Mediterranean climate with cool, wet winters and hot, dry summers. Growing seasons in this analysis are defined by the timing of winter rains which typically occur from October to April, with peak biomass occurring in the spring of each year. Throughout the manuscript growing seasons are referenced by the year in which peak biomass occurred; i.e. the 2013 growing season spanned October 2012 to April 2013.

This site was ideal for evaluating how species composition influences ecosystem responses to drought, because all experimental plots had similar starting soils, aspect, and disturbance history (D. Lipson, unpublished data). The experimental site was a grazed pasture prior to the establishment of the Reserve in 1962, when grazing ceased. The area was then patchily re-colonized by two co-dominant native shrubs: Salvia mellifera (black sage) and Artemisia californica (California sagebrush), with a sparse understory containing both native and exotic herbaceous species. Shrub inter-spaces were dominated by exotic species such as Bromus madritensis (Spanish brome) and Centaurea melitensis (Maltese star thistle), but included native species such as Stipa pulchra (purple needle grass) and Dichelostemma capitatum (blue dicks). Terminology for referring to species origin varies widely (Richardson et al. 2000); throughout this manuscript we refer to all species not native to California as exotic for brevity. We list all species found in the experimental plots, including invasive status of exotic species according to the California Invasive Plant Council, in Supplementary Table 1

Within the site 36 plots were established; six shelter-control plots were uncovered (3 shrub dominated, 3 inter-spaces), while 30 treatment plots were covered with a clear, polycarbonate roof (approximately 1.6 m higher than the ground surface, slanted to allow runoff) which permitted light transmission but intercepted all rainfall during the growing season. Shelter control plots received ambient rainfall, the same amount of rainfall applied to the 100% treatment plots (which had shelter tops). Thus, these plots were not meant as controls for the rainfall treatment, but rather as controls for unintended effects of the experimental infrastructure.

The rainfall shelter tops were 3.5 × 3.5 m2, but all measurements were made in a central 3 × 3 m2 plot (0.5 m buffer from the edge). Half of the plots were centered on mature S. mellifera shrubs with A. californica as a less abundant component of the vegetation (here-after referred to as “shrub understory” plots), while the remaining plots were in shrub inter-spaces dominated by herbaceous vegetation (here-after "open" for brevity). Shrub understory and open plots were paired spatially to reduce bias associated with any unmeasured environmental gradients at the site, at least one meter apart, with paired plots having the same rainfall treatment. The site was nearly flat (little slope); rainfall treatments did not run off of target plots and hence trenching was not necessary between plots. Supplementary Fig1a shows a photograph of the rainfall shelters at the field site, and Supplementary Fig1b shows an aerial view of the spatial arrangement of the plots.

Rainfall shelters were established over treatment plots prior to the 2013 growing season, and roof tops were removed during the dry summer months to permit light penetration and cleaning (and reattached prior to the fall rains). During the growing season, the shelters prevented all ambient rainfall from falling on the plots. All rainfall was collected using a gutter system and held in large storage tanks; experimental rainfall treatments were applied with irrigation on the first dry day following a rain event. Rainfall quantity applied to the experimental treatments was calculated from local weather station data. Rain events exceeding 18 mm were split into multiple application days to reduce surface water run-off from plots.

Three rainfall manipulation treatments were applied during the experiment: 50%, 100% or 150% of ambient rainfall (shrub vs open x three rainfall treatments x five replicates + six non-covered controls = 36 plots). Due to the intense drought which occurred from 2012 to 2016, the 150% rainfall treatment was similar to long-term average precipitation at the site (Fig. 1). Hence, throughout the manuscript we refer to all treatments occurring during the experimental phase of the study as experimental drought treatments (150% treatment approximating "average rainfall," while the 100% and 50% treatments represent moderate and severe drought, respectively).
Fig. 1

Growing season precipitation (October–April) at the experimental site over the last 30 years. Growing season year indicates the year when peak biomass occurs (April). Bar color indicates experimental stage (light gray = historic, dark gray = experimental drought, black = recovery). Rainfall amount experimentally applied during the 4-year manipulative phase of the experiment are shown as circles: yellow = 150% ambient rainfall (average rainfall), orange = 100% (moderate drought),  = 50% (severe drought). The historic 30-year average (1983–2012 growing seasons) at the site is shown as a horizontal line (299 mm)

The rainout shelters were removed following the 2016 growing season, and all plots received ambient rainfall during the 2017 growing season (a higher than average year), to assess the potential for resilience following drought. Growing season (October through April) rainfall totals are shown in Fig. 1 for the last 30 years, including the treatment amounts in the experimental years. Precipitation at the experimental site was measured by the “Fallbrook 5 NE CA US” NOAA weather station for growing season years 2009–2017 (www.wrcc.dri.edu). Prior to the 2009 growing season the "Vista, CA US” station used as the closest station with historic data, 23 km away from the newer Fallbrook station.

Above-ground biomass was collected from a 20 cm × 50 cm area within each plot (0.1 m2) in April, the time of peak growing season biomass, in each of the five growing seasons (four years of rainfall manipulation, and one recovery year where all plots received ambient rainfall). Harvest locations were marked with stakes to prevent re-harvesting in subsequent years. All above-ground biomass of plants rooted in the area was clipped at the soil level—including live biomass as well as recently senesced biomass that reflected current year's growth. All litter, defined as dead plant material from previous growing seasons, was also collected in the harvest area. In the shrub understory, herbaceous biomass harvest areas were positioned to avoid main stems of perennial shrubs, but occasionally contained small shrub seedlings. Biomass in each plot was identified and sorted in the field into three categories: exotic biomass, native biomass, or litter. Biomass samples were dried to a constant mass at 40°C for 72 h in the laboratory and weighed to the nearest 0.01 mg. Woody biomass production of mature shrubs was not quantified. Estimates of above-ground photosynthetic biomass production (integrating across shrub-dominated and herbaceous vegetation) based on normalized difference vegetation index (NDVI) are published elsewhere (Esch 2017), and are consistent with the patterns of herbaceous biomass presented in this analysis.

Biomass is a relevant metric for understanding the impact of species composition on ecosystem-level processes. However, we did not sort biomass to species, and hence we analyzed change in species' abundances in permanent plots to evaluate how life history (annual versus perennial) influenced species' drought response and recovery. In April, for each of the five growing seasons, we visually estimated percent cover of each species within a central 1 × 1 m2 plot marked with permanent metal stakes at the four corners, with a maximum value of 100. When there were overlapping individuals total cover could sum to more than 100. We summed the percent cover of all annual vs perennial species, and separately all native versus exotic species, within each plot. Only understory species were included to mirror the biomass analysis, hence shrub cover was not included in this analysis.

After rainfall manipulations ended, the seed bank was sampled in October 2016 (prior to the onset of the 2017 growing season rains). A 5 cm × 10 cm area of soil was excavated to a depth of 5 cm, along with any litter on the soil surface, adjacent to the area harvested for biomass in April 2016. The collected soil was spread in a thin layer over commercial potting soil in germination trays (30 × 15 × 5 cm) in the UC San Diego Biology Field Station greenhouses, watered daily to keep soil moist and monitored for the identity and abundance of germinating seedlings for a period of 6 months. The total number of germinating seeds was calculated as a measure of seed bank abundance for native versus exotic species. A summary of the experimental design is shown in Fig. 2, and additional detailed methods are described in Esch (2017).
Fig. 2

Conceptual figure of the experimental design. Plots were covered with clear rainout shelters (indicated by light blue color) during years with rainfall manipulation (drought stage, 2013–2016). During the drought stage plots were experimentally given 50%, 100%, or 150% of ambient rainfall. In the recovery stage, all plots were uncovered and exposed to ambient rainfall (white). In all years, the herbaceous biomass from the shrub understory and open plots was harvested and sorted into exotic biomass (yellow), native biomass (green), or litter (white)

Statistical analyses

All analyses were conducted in R version 3.2.3 (R Core Development Team 2016). Initial analyses of biomass and cover in the 2013–2016 growing seasons were conducted with general linear mixed models using the lme call in the package nlme (Pinheiro et al, 2013). These models included growing season rainfall (continuous), species origin, and their interactions; plot was included as a random factor to account for the repeated-measures design across years. The influence of life-history differences was evaluated in a separate model, where percent cover was predicted by rainfall (continuous), life history (annual versus perennial), and their interactions (biomass was only separated by species origin, and hence could not be analyzed for life-history differences). Data from the shrub understory versus open plots were analyzed in separate models. As described in the results below, the initial analysis revealed that biomass was influenced by species origin x rainfall interactions. Hence, native and exotic biomass were subsequently analyzed in separate models to facilitate the interpretation of complex statistical interactions. Significance for each factor was evaluated with Type II Wald chi-square tests using the Anova function in the car package (Fox and Weisberg 2011).

To evaluate the influence of antecedent effects (litter and previous year rainfall) in addition to the current year’s rainfall, we used a model averaging approach (Grueber et al. 2011). This approach is useful in cases such as ours where wished to evaluate the relative weight of models with and without antecedent effects. We used Akaike’s Information criteria (AIC) to determine the support for each individual model in a set of models. The potential input variables were (i) litter mass (ii) current year's rainfall and (iii) previous year’s rainfall. To generate a sub-model set, we first fit a global general linear mixed model (Bates et al. 2014) containing all three variables. We standardized the input variables using the stdize function in the MuMIn package, so that the parameter estimates could compared among factors with different measurement scales after model averaging (Gelman 2008). Next, we used the dredge function in the MuMIn package to consider all combinations of the global model (Bartoń 2016). We used get.models to select a subset of those models which varied by AIC < 4, performed model averaging on these top submodels, and reported the conditional estimate results (following the procedure of Grueber et al. 2011).

In keeping with the structure of the prior analyses, the number of individuals emerging from the seed bank was analyzed separately between shrub understory and open plots. Here, we used a linear model including species origin and experimental drought treatment (as 50%, 100%, or 150% of ambient rainfall, continuous) as fixed factors.

Resilience of biomass in the recovery year (2017) was analyzed using a set of steps similar to the drought response analyses. Biomass in 2017 was first analyzed in a linear model where species origin, previous drought treatment (50%, 100%, or 150% of ambient rainfall) and their interaction were included as fixed factors, with separate models for shrub understory versus open plots. Given species origin by treatment interactions (see “Results”), the subsequent analysis of antecedent effects was conducted separately for native versus exotic biomass to enable a mechanistic evaluation of the statistical interactions. Again the role of antecedent factors in drought recovery (previous year’s rainfall, standing litter, and number of germinants in the 2016 seedbank) was assessed with a model averaging approach, using the same methods previously described for the analysis of drought response.

Results

Site community composition and shelter controls

A total of 43 species were recorded in our analyses of percent cover across all plots and years in this experiment (Supplementary Table 1). Seventeen of the 24 native species were perennial, while 15 of the 19 species of exotic origin were annual (Table 1), reflecting a key difference in the dominant life-strategies of native versus exotic species in this system. In contrast there were not major morphological differences between origin groups; both native and exotic cover was approximately equally comprised of grasses and forbs (data not shown). Exotic species dominated the herbaceous community in this system, representing approximately 3/4 of both cover and biomass (Table 1).
Table 1

Summary of community composition averaged across all experimental plots during the 5-year experiment, including species origin (native or exotic to California), life history (annual or perennial), mean total cover (summed cover within each group, can be greater than 100%), fraction of total cover (proportion of total cover made up by each group, sums to 1), species richness, and mean total herbaceous biomass (note that biomass was not sorted into annual versus perennial species, so this represents total exotic or native biomass, averaged across all plots)

Origin

Life history

Mean total % Cover

Fraction of total cover

Species richness

Mean herbaceous biomass (g/m2)

Exotic

Annual

97.9

0.72

15

31.71

Perennial

7.4

0.05

4

 

Native

Annual

5.7

0.04

7

7.50

Perennial

24.6

0.18

17

 

Biomass of native and exotic species were compared between the unsheltered controls and the sheltered 100% plots, which both received the same amount of rainfall. There was no difference between the control and 100% rainfall plots for native biomass, nor for exotic biomass in the shrub understory, but in the open there was greater exotic biomass in the unsheltered controls than in the 100% rainfall plots (Supplementary Fig. 2). All rainfall treatment plots had shelter tops, and hence relative responses of biomass to the drought treatments can be compared across treatments. However, we proceed with the caveat that absolute values of exotic biomass presented in the manuscript are lower than would be observed in the absence of the shelters.

Drought response and recovery

The rainfall manipulation experiment corresponded with the lowest period of rainfall on record for our site, and our lowest rainfall treatments represented a severe 4-year drought (Fig. 1). Consistent with expectations, plant growth (peak season biomass) declined in response to our drought treatments, but with differences between native and exotic species which depended on context (shrub understory versus open vegetation). In the open plots, exotic biomass declined more with drought than native biomass (significant Rainfall × Origin interaction, Table 2). However, in the shrub understory exotic and native biomass had similar declines in response to drought (significant main effect of Rainfall, Table 2). Overall, there was lower herbaceous biomass in the shrub understory than in the open areas, and exotic and native biomass were more equally represented in the shrub understory, while exotic biomass dominated in the open herbaceous plots (compare Fig. 3 A vs B).
Table 2

Statistics from linear models evaluating how species origin (native/exotic) and rainfall influenced herbaceous biomass during four years of experimental rainfall manipulation (response during 2013–2016 growing seasons, top), and in the recovery year (2017 growing season bottom)

Timeframe

Context

 

χ 2

df

p value

Response to drought (2013–2016)

Open

Rainfall

56.6

1

< 0.001

Origin

41.6

1

< 0.001

Rainfall:Origin

33.1

1

< 0.001

Shrub understory

Rainfall

6.69

1

0.009

Origin

2.88

1

0.09a

Rainfall:Origin

0.29

1

0.59

Open

Treatment

0.60

1

0.45

Recovery (2017)

Origin

53.1

1

< 0.001

Treatment:Origin

5.05

1

0.033

Shrub understory

Treatment

0.73

1

0.40

Origin

2.07

1

0.16

Treatment:Origin

0.37

1

0.54

Rainfall is treated as a continuous variable in these analyses, because the experimental drought treatments resulted in different total growing season rainfall amounts among years. The multi-year analysis of response to drought included plot as a random repeated factor, and hence the mixed model analysis of deviance statistic is presented as χ2. Note that in the recovery year (2017) all plots received the same rainfall, and hence the model term indicating the experimental rainfall treatment in the prior year is abbreviated as “Treatment.” Significant terms are highlighted in bold

aIndicates marginal significance

Fig. 3

Herbaceous biomass versus total growing season rainfall during the experimental drought (2013–2016, a, b) and recovery (2017, c, d) stages. Biomass was sorted to exotic (yellow) or native (green) origin, and quantified in both in open plots (a, b) and in the shrub understory (c, d). Means (n = 5) and ±1 SE of the mean are shown. Shape in the recovery year (c, d) indicates previous rainfall treatment: square =50% of ambient rainfall (severe drought), circle = 100% (moderate drought), triangle = 150% (average rainfall). Note different range of values on the vertical axes

During the recovery year (2017), all plots received some of the highest rainfall on record at the experimental site (Fig. 1). In the open plots, native and exotic biomass differed in the dynamics of their recovery following the secession of drought treatments (Treatment x Origin interaction, Table 2). Open plots that had previously experienced drought treatments had lower native biomass, but higher exotic biomass, compared to plots that had received the highest rainfall treatment (Fig. 3 C). In contrast, in the shrub understory, the prior rainfall treatments did not influence growth of native or exotic species in the recovery year (Fig. 3 D).

The drought response and recovery of percent cover by species origin were similar to those for biomass (Supplementary Fig. 3), and for the sake of brevity are not discussed in the main text.

The influence of antecedent factors in drought response

During the drought, current rainfall was the most important predictor of biomass, and there were few significant influences of antecedent factors (Table 3); for instance, previous year's rainfall did not influence native or exotic biomass. However, in the open plots, standing litter from the previous growing season had a positive effect on exotic biomass during drought (Table 3). In the shrub understory, both native and exotic biomass had a positive relationship with current rainfall, but no antecedent factors were important predictors.
Table 3

Biomass responses to drought treatments (2012–2016) as influenced by both concurrent and antecedent factors

Context

Biomass

Estimate

SE

z value

p value

Importance value

Type

Model term

Open

Native

Rainfall

1.8E−02

4.1E−03

4.42

< 0.001

1

Previous Rainfall

 4.7E03

3.0E03

1.53

0.13

0.52

Litter

1.4E02

1.3E02

1.13

0.26

0.38

Exotic

Rainfall

9.4E−02

1.3E−02

7.09

<0.001

1

Previous Rainfall

 7.3E03

1.1E02

0.67

0.50

0.28

Litter

1.2E−01

4.6E−02

2.57

0.01

1

Shrub understory

Native

Rainfall

4.7E−03

1.8E−03

2.57

0.01

0.92

Previous Rainfall

1.8E-03

1.603

1.17

0.24

0.41

Litter

 2.7E-03

2.3E03

1.16

0.25

0.36

Exotic

Rainfall

2.2E-03

1.0E−03

2.15

0.03

0.89

Previous Rainfall

 3.7E04

8.7E04

0.42

0.67

0.18

Litter

8.6E04

1.4E03

0.62

0.53

0.2

This table shows the output from a model averaging analysis evaluating the relative importance of growing season rainfall (Rainfall), growing season rainfall in the previous year (Previous Rainfall), and Litter for predicting native and exotic herbaceous biomass in open plots versus in the shrub understory. SE=Adjusted standard error from the model output

Significant treatment effects are highlighted in bold

Abundance of native and exotic species in the seedbank following drought

Of the 1095 seedlings which emerged in the greenhouse from our field-collected soils, we were able to identify just over half to the species-level and hence identify the seedling as either of native or exotic origin (566 seedlings). Species identified from the seedbank are indicated in Table S1, and accounted for 70% of total species cover observed in the experiment. Early mortality of emerging seedlings limited our ability to identify remaining seedlings. There was a greater mean number of seedlings germinating in soils collected from open plots compared to soils collected under the shrub understory (Fig. 4), and exotic seedlings dominated the open plots while native species were more common in the shrub understory (marginally significant main effects of Origin, Table 4). In soils from open plots, there was a significant Origin × Treatment interaction (Table 4), whereby exotic seedling abundance was lowest in plots that had experienced the drought treatments, but native seedling abundance did not vary with treatment. A visual examination of the total number of germinating individuals (identified to species plus unidentified) reveals very similar overall patterns of seedbank responses to prior rainfall treatments (Supplementary Fig. 4), suggesting the dynamics of the identified species reflect the overall seedbank responses, with the caveat that individual species dynamics in the unidentified pool could have differed significantly from the identified species.
Fig. 4

Number of germinating individuals from soils collected in October 2016, following 4 years of experimental drought treatments. Only individuals that could be identified to the species level are included, according to the species origin as exotic (yellow) or native (green). Soils were collected from both the shrub understory (left panel) and open plots (right panel)

Table 4

Statistics from linear models evaluating how native versus exotic origin (Origin) and experimental drought treatments (Treatment) influenced variance in the abundance of germinating seeds from the seedbank, for soils collected in October 2016, following four growing seasons

Context

Factor

F value

df

p value

Open

Treatment

2.93

1

0.13

Origin

2.53

1

0.08

Treatment:Origin

3.32

1

0.03

Shrub understory

Treatment

2.35

1

0.15

Origin

2.24

1

0.09

Treatment:Origin

5.25

1

0.25

Significant terms are indicated in bold font

The influence of antecedent factors in drought recovery

As described previously, exotic and native biomass in the open plots had opposing relationships with prior rainfall treatments in recovery year, as evidenced by a Rainfall x Origin interaction (Table 2, Fig. 2c). To investigate the influence of antecedent factors, native versus exotic biomass were analyzed separately using a model averaging approach including prior rainfall treatment, litter, and seedbank abundance as continuous variables. In the open plots, prior rainfall treatment had a significant positive effect on native biomass, but a negative effect on exotic biomass (Table 5). Litter mass remaining in plots from the prior year's plant growth had a significant, positive effect on native biomass in both open plots and in the shrub understory (Table 5).
Table 5

Recovery following drought (2017) as influenced by both concurrent and antecedent factors

Context

Biomass

Estimate

SE

z value

p value

Importance value

Type

Model term

Open

Native

Litter

6.3E01

1.7E01

3.66

<0.001

1

Treatment

7.9E02

1.4E02

5.75

<0.001

1

Seedbank

1.5E−02

3.0E−02

0.49

0.63

0.27

Exotic

Litter

7.9E−01

4.5E−01

1.75

0.08

0.62

Treatment

− 1.6E01

3.8E02

4.31

<0.001

1

Seedbank

1.2E−01

7.9E-02

1.49

0.14

0.51

Shrub understory

Native

Litter

2.3E02

7.0E03

3.25

0.001

1

Treatment

4.7E−03

5.5E−03

0.85

0.39

0.31

Seedbank

− 7.1E−02

5.5E−02

1.29

0.19

0.43

Exotic

Litter

3.4E−03

7.5E−03

0.46

0.65

0.18

Treatment

− 7.7E−04

5.8E−03

0.13

0.89

0.16

Seedbank

− 1.1E−02

5.3E−02

0.20

0.84

0.16

This table shows the output from a model averaging analysis evaluating the roles of 2013–2016 growing season experimental rainfall treatment (Treatment), grams/m2 standing dead litter (Litter), and the number of germinating seeds in the seedbank in October 2016 (seedbank) for predicting native and exotic herbaceous biomass following drought in open plots versus in the shrub understory

Significant terms are indicated in bold font

The influence of life history on drought response and recovery

Total percent cover also declined in response to the drought treatments, but differed between annual and perennial species and depended on context (shrub understory versus open vegetation). In the open plots, annual species’ percent cover declined more with drought than perennial species (significant Rainfall x Life history interaction, Table 6). However, in the shrub understory neither life history nor rainfall treatment explained significant variation in percent cover (Table 6). Overall, there was lower cover of annual species in the shrub understory than in the open plots, and cover of perennial species was higher in the shrub understory than in the open plots (compare Fig. 5a vs b). These responses of annual versus perennial cover to drought are nearly identical to responses of exotic versus native cover (Supplementary Fig3), and biomass (Fig. 2a, b), demonstrating the important role of life history for understanding the drought responses of native versus exotic species in this system.
Table 6

Statistics from linear models evaluating how species life history (annual/perennial) and rainfall influenced percent cover during 4 years of experimental rainfall manipulation (response during 2013–2016 growing seasons, top), and in the recovery year (2017 growing season bottom)

Timeframe

Context

Model term

Sum Sq

Df

F value

p value

Response to drought (2013–2016)

Open

Rainfall

5093

1

47.57

< 0.001

Life history

3446

1

32.19

< 0.001

Rainfall:Life

2273

1

21.24

< 0.001

Shrub understory

Rainfall

25.02

1

1.55

0.23

Life history

1.77

1

0.11

0.74

Rainfall:Life

59.42

1

3.68

0.07a

Recovery (2017)

Open

Treatment

163.1

1

95.3

0.01

Life history

1926

1

1125

< 0.001

Treatment:Life

5.34

1

3.12

0.22

Shrub understory

Treatment

0.64

1

0.36

0.61

Life history

0.011

1

0.01

0.94

Treatment:Life

0.36

1

0.20

0.69

Rainfall is treated as a continuous variable in these analyses, because the experimental drought treatments resulted in different total growing season rainfall amounts among years. Note that in the recovery year (2017) all plots received the same rainfall, and hence the model term indicating the experimental rainfall treatment in the prior year is abbreviated as Treatment

Significant terms are highlighted in bold

aIndicates marginal significance

Fig. 5

Species percent cover in response to precipitation amount during the experimental drought (2013–2016, a, b) and recovery year (2017, c, d). Color indicates summed percent cover of annual (blue) or perennial (red) species, in open plots (A, B) and in the shrub understory (c, d). Means (n = 5) and ±1 SE of the mean are shown

However, during the recovery year (2017), the dynamics of annual versus perennial cover were not predicted by differences in species origin. In the open plots, annual cover was greater than perennial cover overall (main effect of Life history, Table 6), but the cover of both annual and perennial species was negatively impacted by prior drought severity (main effect of Treatment Table 6, Fig. 5c). In contrast, the responses of exotic versus native biomass to drought from 2013–2016 were nearly identical to the cover responses of annual versus perennial species, respectively (compare Figs. 2a and 5a). In the shrub understory plots, neither rainfall nor life history had a significant effect on species percent cover, again highlighting how context influenced species recovery from drought.

Discussion

Temporal dynamics have long been at the core of community ecology theory, both for predicting species coexistence, and changes in community composition in response to variation in the environment (e.g. Holling 1973; Sousa 1984; Chesson 1994). These theories are ever more relevant as ecologists seek to predict how ecosystems will respond to an increasingly variable global climate (Collins et al. 2013), and continued introductions of exotic species (Bellard et al. 2013; Seebens et al. 2015). The response of an ecological system to a disturbance, such as severe drought, is often defined in terms of resistance and resilience—the magnitude of the response compared to baseline conditions and how quickly the system recovers to the pre-disturbance state, respectively (Holling 1973; Westman 1978). In this experiment, we found that native biomass was more resistant to drought, but exotic biomass was more resilient. The response of exotic biomass to drought was strongly driven by life history, with annual cover also responding negatively to drought. The drought recovery of exotic species was not explained by life history, however, because annual and perennial species had similar patterns of drought recovery. We also found that context played an important role, where drought response and recovery were more influenced by rainfall in open plots than in the shrub understory. Antecedent factors were also more important in open plots, and sometimes differed in their influence on native versus exotic species. For instance, in the recovery year, prior rainfall treatment was a positive predictor of native biomass, but a negative predictor of exotic biomass, only in the open plots. In contrast, litter was an antecedent factor that tended to influence native and exotic biomass in the same way, albeit at different times. Litter from the prior year's growth had a positive effect on exotic biomass in the open plots during the response year, and on native biomass in the recovery year regardless of context. These findings will be detailed in the remainder of the discussion.

In the open plots, exotic biomass and abundance (as estimated by cover) declined more steeply than native biomass in response to drought, consistent with the findings of recent meta-analyses of experimental rainfall reduction (Sorte et al. 2013; Liu et al. 2017). Our result is also consistent with other experiments and observations showing a reduction in the abundance of exotic herbaceous species in response to drought in California (Copeland et al. 2016). However, in the first high-rainfall year following drought, exotic biomass rebounded from nearly absent to the highest levels observed during the experiment. It is possible that exotic seeds remaining in the soil following drought were able to produce large individuals in the high-rainfall year of 2017, a kind of demographic inertia that is especially beneficial to species in variable environments. Drought can increase resources (including space) by reducing the growth of resident vegetation, enabling fast-growing annual, exotic species to quickly establish when favorable growing conditions return (Jiménez et al. 2011; Diez et al. 2012). Consistent with this prediction, Kimball et al. (2014) found that extreme drought following fire prevented native shrub re-establishment, and ultimately facilitated conversion of coastal sage scrub to an herbaceous, invaded vegetation type.

Our finding that exotic biomass was more sensitive to drought than native biomass is strongly associated with the dominant life histories of the two groups. When we analyzed the percent cover responses of annual versus perennial species, we saw that their responses were nearly identical to those of exotic versus native species, respectively. Our results are consistent with prior work showing that across herbaceous-dominated ecosystems in the United States, annual species are more sensitive to interannual variation in rainfall than perennial species (Cleland et al. 2013). This also suggests that our results are likely to be relevant for other Mediterranean-type ecosystems, where annuals dominate exotic species pools (Funk et al. 2016).

We found mixed responses of biomass to antecedent rainfall; during the years of drought, prior rainfall did not influence biomass, but in the recovery year native and exotic biomass had opposing relationships with prior rainfall. A recent meta-analysis found a signal of lagged rainfall effects across an aridity gradient from deserts to mesic grasslands, where prior year's rainfall was proportionally more important for predicting production in wet sites (Sala et al. 2012). Therefore, our site may be on the dry end of a gradient where prior year's rainfall is less important than current year's rainfall for influencing plant growth. However, their analysis did not distinguish among functional groups, which can respond differently to drought. For instance, an experimentally severe drought in a tall-grass prairie reduced productivity overall, but impacted dominant forbs more than dominant grasses (Hoover et al. 2014). The grasses rebounded following drought due to a strong demographic effect, resulting in a persistent shift in community composition. The Sala et al. (2012) analysis also did not include other antecedent factors such as litter accumulation, which could be correlated with prior year's rainfall but be more mechanistically predictive.

Standing litter had a positive effect on exotic biomass during the drought, and on native biomass in the recovery year. Although litter tended to have a similar influence on native and exotic biomass in our study, litter is likely to have differing influences on growth and abundance among species. For instance, Dudney et al. (2017) found that high annual rainfall increased litter in the following year, suppressing forb growth but increasing grasses, resulting in stronger lagged precipitation effects on forbs. Litter can modify germination rates (Reynolds et al. 2001), reduce light (Foster and Gross 1998), protect seedlings from herbivory (Facelli and Pickett 1991), and increase soil moisture (Wolkovich et al. 2010). Hence, litter accumulation can be a key mechanism linking previous year’s rainfall to shifts in production and species composition in the following years. A recent meta-analysis found a positive relationship between litter and new biomass production across common garden experiments, but a neutral relationship in field experiments (Loydi et al. 2013). Interestingly, field studies in dry sites tend to find a positive relationship between litter and new biomass production (Boeken and Orenstein 2001; Eckstein et al. 2012), suggesting that the influence of litter on plant growth is likely context dependent. At our semi-arid site shade, protection from herbivory or increased soil moisture may underlie the positive relationship between litter and biomass we observed in this experiment.

In the open plots, the number of exotic seedlings emerging from the seedbank was positively associated with prior rainfall treatments, and hence the seedbank could not explain the dramatic recovery of exotic biomass in plots that had experienced the drought treatments. In contrast in the shrub understory, we found a trend towards greater abundance of native germinants in plots that had experienced drought compared to plots that had experienced average rainfall, even though native biomass showed the greatest recovery following the average rainfall treatments. This potentially suggests that native species, at least in the shrub understory, did not break dormancy during the prolonged drought, and seeds accumulated over this time. La Forgia et al. (2018) found accumulation of dormant seeds during drought of native forbs but not exotic grasses, in a more mesic system in Northern California. Capacity for persistence in soil is common in species with variable reproductive success across years (Pake and Venable 1996), and persistent seed banks are common in the Mediterranean-climate regions of California that experience high-interannual variability in rainfall (Parker and Kelly 1989). Few studies have compared the dynamics of native versus exotic seedbanks (but see Faist et al. 2013; Gioria and Pyšek 2016). However, we might expect native species to have longer lived seedbanks than exotic species in this system for two reasons; first, exotic species often germinate faster, and across a wider array of environmental conditions, compared with native species (Chrobock et al. 2011; Wilsey et al. 2011; Wainwright and Cleland 2013), and hence would not be expected to leave a persistent seed bank. Second, exotic grasses in California have larger seeds than native grasses (Sandel and Dangremond 2012), and large seeded species have lower persistence in the seedbank (Rees 1994). However, the generality of this finding will depend on differences in seed size between native and exotic species, and are likely to vary across systems because of observed latitudinal and regional variation in seed size (Moles et al. 2008). We should also note the important caveat that nearly half of the emerging seedlings suffered mortality before they could be identified, a common issue in seedbank studies that limits our ability to say with certainty how seedbank abundance is related to species percent cover observed at the end of the growing season.

Given that neither litter nor the seedbank could mechanistically predict the positive influence of prior drought on exotic biomass in the recovery year, what mechanism might have been responsible? One possibility is nitrogen accumulation due to reduced plant uptake or decreased leaching losses during the drought; a recent meta-analysis of rainfall reduction experiments showed that inorganic nitrogen, especially ammonium, can accumulate under dry conditions (Homyak et al. 2017). Consistent with this hypothesis, inorganic nitrogen accumulated more over time in our drought treatments than the treatments receiving higher rainfall (Castro 2018). This suggests that the increase in exotic biomass following severe drought in our experiment may have been caused by high-nutrient availability. For instance, nitrogen enrichment often promotes invasion by fast-growing exotic species (e.g. Huenneke et al. 1990; Bobbink et al. 2010), especially annuals (Suding et al. 2005). A study in Northern California found that exotic growth was only promoted by high rainfall on fertile sites and in the absence of competitors (Eskelinen and Harrison 2014), highlighting how exotic growth in our experiment could be promoted by high rainfall after a period of drought during which nitrogen accumulated in soil. While not directly testable in this experimental design, drought may have also altered competitive interactions between native and exotic species; shifting species interactions (Suttle et al. 2007) could cause shifts in community-level responses to environmental change, such as the ones observed in this study.

One of the most striking findings of this study is how dynamics of biomass and cover in the shaded understory of the native shrubs differed from the responses in the open, herbaceous-dominated plots. In the understory, native and exotic biomass declined to the same degree with drought, and also recovered to the same extent in 2017. Biomass in the shrub understory was made up equally of native and exotic species, and exotic biomass was markedly lower than in the open plots; potentially reflecting reduced levels of competition for soil resources between native and exotic species in the low-light conditions of the shrub understory (Soliveres et al. 2011). This suggests that dynamics of resistance and resilience in response to drought may be dampened in stressful or low-resource environments, and is a promising avenue for future research. This also suggests that as shrub-dominated habitats are increasingly converted to exotic annual grassland, as a result of global changes such as accelerating fire regimes, nitrogen deposition (Talluto and Suding 2008) and drought (Kimball et al. 2014), we could expect to see a shift in the dynamics of drought response and recovery in these systems, whereby ecosystem biomass production dominated by exotic annual species is increasingly sensitive to interannual variation in rainfall.

In conclusion, we found that exotic biomass declined steeply in response to our imposed severe drought, but recovered in the first year of high rainfall, to levels higher than in experimental treatments that had maintained close to average rainfall at the site. This result is strongly associated with the life histories of the dominant exotic (mostly annual) versus native (mostly perennial) species in this system. Our results are likely to extend to other Mediterranean-type ecosystems where annual species dominate the exotic species pool (Funk et al. 2016). Our results also suggest that predicted increases in interannual rainfall variability (drought followed by high rainfall) could further promote growth of exotic species in this system, particularly in open areas where shrub canopy is sparse. We also found that antecedent factors, specifically seedbank abundance and litter from the prior year's production, differed in their influence on native versus exotic biomass depending on spatial context (shrub understory versus open plots for seed bank) or temporal context (during or following drought for litter). Together these results demonstrate how temporal variability in rainfall differentially influences growth and abundance of native versus exotic species, with important ramifications for ecosystem dynamics and biodiversity in this diverse region.

Notes

Acknowledgements

This work was supported by National Science Foundation (NSF) Division of Environmental Biology Grants to E.E.C. (DEB-1154082) and to D.L. (DEB-1153958), and NSF Graduate Research Fellowships to E.H.E (DGE-1144086) and C.E.P. (DGE-16540112). A California Native Plant Society Grant to E.H.E. also helped support this research. We thank Rachel Abbott, Andrew Heath, Christopher Kopp, and Elizabeth Premo for help in maintaining the field experiment. This work was performed at the San Diego State University’s Santa Margarita Ecological Reserve and we thank Pablo Bryant for site access and maintenance.

Supplementary material

11258_2019_912_MOESM1_ESM.docx (5.9 mb)
Supplementary file1 (DOCX 6085 kb)

References

  1. Anderson JE, Inouye RS (2001) Landscape-scale changes in plant species abundance and biodiversity of a sagebrush steppe over 45 years. Ecol Mon 71:531–556.  https://doi.org/10.2307/3100035 CrossRefGoogle Scholar
  2. Ashbacher AC, Cleland EE (2015) Native and exotic plant species show differential growth but similar functional trait responses to experimental rainfall. Ecosphere 6:1–14.  https://doi.org/10.1890/ES15-00059.1 CrossRefGoogle Scholar
  3. Balch JK, Bradley BA, D'Antonio CM, Gómez-Dans J (2013) Introduced annual grass increases regional fire activity across the arid western USA (1980–2009). Glob Change Biol 19:173–183.  https://doi.org/10.1111/gcb.12046 CrossRefGoogle Scholar
  4. Bartoń K (2016) MuMIn: Multi-Model Inference. R package version 1(15):7Google Scholar
  5. Bates D, Maechler M, Bolker B, Walker S (2014) lme4: Linear mixed-effects models using Eigen and S4.Google Scholar
  6. Bellard C, Thuiller W, Leroy B, Genovesi P, Bakkenes M, Courchamp F (2013) Will climate change promote future invasions? Glob Change Biol 19:3740–3748.  https://doi.org/10.1111/gcb.12344 CrossRefGoogle Scholar
  7. Berg N, Hall A (2015) Increased interannual precipitation extremes over California under climate change. J Clim 28:6324–6334.  https://doi.org/10.1175/JCLI-D-14-00624.1 CrossRefGoogle Scholar
  8. Bobbink R, Hicks K, Galloway J, Spranger T, Alkemade R, Ashmore M, Bustamante M, Cinderby S, Davidson E, Dentener F, Emmett B, Erisman J-W, Fenn M, Gilliam F, Nordin A, Pardo L, De Vries W (2010) Global assessment of nitrogen deposition effects on terrestrial plant diversity: a synthesis. Ecol Appl 20:30–59.  https://doi.org/10.1890/08-1140.1 CrossRefGoogle Scholar
  9. Boeken B, Orenstein D (2001) The effect of plant litter on ecosystem properties in a Mediterranean semi-arid shrubland J Veg Sci 12:825–832Google Scholar
  10. Bradley BA (2009) Regional analysis of the impacts of climate change on cheatgrass invasion shows potential risk and opportunity. Glob Change Biol 15:196–208.  https://doi.org/10.1111/j.1365-2486.2008.01709.x CrossRefGoogle Scholar
  11. Castro, S. (2018) Impacts of altered rainfall and invasive plants on soil microbial communities in southern California shrublands. Dissertation, San Diego State UniversityGoogle Scholar
  12. Cavaleri MA, Sack L (2010) Comparative water use of native and invasive plants at multiple scales: a global meta-analysis. Ecology 91:2705–2715.  https://doi.org/10.1890/09-0582.1 PubMedCrossRefGoogle Scholar
  13. Cayan DR, Maurer EP, Dettinger MD, Tyree M, Hayhoe K (2008) Climate change scenarios for the California region. Clim Change 87:21–42.  https://doi.org/10.1007/s10584-007-9377-6 CrossRefGoogle Scholar
  14. Chapin FS III (1980) The mineral nutrition of wild plants. Annu Rev Ecol Syst 11:233–260.  https://doi.org/10.1146/annurev.es.11.110180.001313 CrossRefGoogle Scholar
  15. Chesson P (1994) Multispecies competition in variable environments. Theor Pop Biol 45:227–276.  https://doi.org/10.1006/tpbi.1994.1013 CrossRefGoogle Scholar
  16. Chrobock T, Kempel A, Fischer M, van Kleunen M (2011) Introduction bias: cultivated alien plant species germinate faster and more abundantly than native species in Switzerland. Basic App Ecol 12:244–250.  https://doi.org/10.1016/j.baae.2011.03.001 CrossRefGoogle Scholar
  17. Cincotta RP, Wisnewski J, Engelman R (2000) Human population in the biodiversity hotspots. Nature 404:990–992.  https://doi.org/10.1038/35010105 PubMedCrossRefGoogle Scholar
  18. Cleland EE, Collins SL, Dickson TL, Farrer EC, Gross KL, Gherardi LA, Hallett LM, Hobbs RJ, Hsu JS, Turnbull L, Suding KN (2013) Sensitivity of grassland plant community composition to spatial vs. temporal variation in precipitation. Ecology 94:1687–1696.  https://doi.org/10.1890/12-1006.1 PubMedCrossRefGoogle Scholar
  19. Cleland EE, Funk JL, Allen EB (2016) Coastal sage scrub. In: Mooney H, Zavaleta E (eds) Ecosystems of California. University of California Press, Berkley, pp 429–448Google Scholar
  20. Collins M, Knutti R, Arblaster J, Dufrense J-L, Fichefet T, Friedlingstein P, Gao X, Gutowkski WJ, Johns T, Krinner G, Shongwe M, Tebaldi C, Weaver AJ, Wehner M (2013) Long-term climate change: projections, commitments and irreversibility. In: Stocker TF, Qin D, Plattner G-K, Tignor M, Allen SK, Boschung J, Nauels A, Xia Y, Bex V, Midgley PM (eds) Climate change 2013: the physical science basis. Contribution of Working Group I to the Fifth Assessment Report of the Intergovernmental Panel on Climate Change. Cambridge University Press, Cambridge, United Kingdom and New York, NY, USA.Google Scholar
  21. Concilio AL, Nippert JB, Ehrenfeucht S, Cherwin K, Seastedt TR (2016) Imposing antecedent global change conditions rapidly alters plant community composition in a mixed-grass prairie. Oecologia 182:899–911.  https://doi.org/10.1007/s00442-016-3684-4 PubMedCrossRefGoogle Scholar
  22. Copeland SM, Harrison SP, Latimer AM, Damschen EI, Eskelinen AM, Fernandez-Going B, Spasojevic MJ, Anacker BL, Thorne JH (2016) Ecological effects of extreme drought on Californian herbaceous plant communities. Ecol Monogr 86:295–311.  https://doi.org/10.1002/ecm.1218 CrossRefGoogle Scholar
  23. Cox RD, Allen EB (2008) Composition of soil seed banks in southern California coastal sage scrub and adjacent exotic grassland. Plant Ecol 198:37–46.  https://doi.org/10.1007/s11258-007-9383-9 CrossRefGoogle Scholar
  24. Daehler CC (2003) Performance comparisons of co-occurring native and alien invasive plants: implications for conservation and restoration. Ann Rev Ecol Sys 34:183–211.  https://doi.org/10.1146/annurev.ecolsys.34.011802.132403 CrossRefGoogle Scholar
  25. D'Antonio CM, Vitousek PM (1992) Biological invasions by exotic grasses, the grass/fire cycle, and global change. Ann Rev Ecol Evol Sys 23:63–87.  https://doi.org/10.1146/annurev.es.23.110192.000431 CrossRefGoogle Scholar
  26. Dennison PE, Brewer SC, Arnold JD, Moritz MA (2014G) Large wildfire trends in the western United States, 1984–2011. Geophys Res Lett 41:2928–2933.  https://doi.org/10.1002/2014GL059576 CrossRefGoogle Scholar
  27. Diez JM, D’Antonio CM, Dukes JS, Grosholz ED, Olden JD, Sorte CJ, Blumenthal DM, Bradley BA, Early R, Ibáñez I, Jones SJ, Lawler JJ, Miller LP (2012) Will extreme climatic events facilitate biological invasions? Front Ecol Env 10:249–257.  https://doi.org/10.1890/110137 CrossRefGoogle Scholar
  28. Diffenbaugh NS, Giorgi F (2012) Climate change hotspots in the CMIP5 global climate model ensemble. Climatic Change 114:813–822.  https://doi.org/10.1007/s10584-012-0570-x PubMedPubMedCentralCrossRefGoogle Scholar
  29. Diffenbaugh NS, Swain DL, Touma D (2015) Anthropogenic warming has increased drought risk in California. PNAS 112:3931–3936.  https://doi.org/10.1073/pnas.1422385112 PubMedCrossRefGoogle Scholar
  30. Dudney J, Hallett LM, Larios L, Farrer EC, Spotswood EN, Stein C, Suding KN (2017) Lagging behind: have we overlooked previous-year rainfall effects in annual grasslands? J Ecol 105:484–495.  https://doi.org/10.1111/1365-2745.12671 CrossRefGoogle Scholar
  31. Eckstein RL, Ruch D, Otte A, Donath TW (2012) Invasibility of a nutrient-poor pasture through resident and non-resident herbs is controlled by litter, gap size and propagule pressure. PLoS ONE 7:e41887.  https://doi.org/10.1371/journal.pone.0041887 PubMedPubMedCentralCrossRefGoogle Scholar
  32. Ehrenfeld JG (2010) Ecosystem consequences of biological invasions. Ann Rev Ecol Evol Sys 41:59–80.  https://doi.org/10.1146/annurev-ecolsys-102209-144650 CrossRefGoogle Scholar
  33. Esch EH (2017) Invasion increases ecosystem sensitivity to drought in Southern California. Dissertation. University of California, San DiegoGoogle Scholar
  34. Eskelinen A, Harrison S (2014) Exotic plant invasions under enhanced rainfall are constrained by soil nutrients and competition. Ecology 95:682–692.  https://doi.org/10.1890/13-0288.1 PubMedCrossRefGoogle Scholar
  35. Facelli JM, Pickett STA (1991) Plant litter: Its dynamics and effects on plant community structure. Bot Rev 57:1–32.  https://doi.org/10.1007/BF02858763 CrossRefGoogle Scholar
  36. Faist AM, Ferrenberg S, Collinge SK (2013) Banking on the past: seed banks as a reservoir for rare and native species in restored vernal pools. AoB Plants 5:plt043.  https://doi.org/10.1093/aobpla/plt043 PubMedCentralCrossRefGoogle Scholar
  37. Foster BL, Gross KL (1998) Species richness in a successional grassland: effects of nitrogen enrichment and plant litter. Ecology 79:2593–2602.  https://doi.org/10.1890/0012-9658(1998)079[2593:SRIASG]2.0.CO;2 CrossRefGoogle Scholar
  38. Fox J, Weisberg S (2011) An R companion to applied regression. SAGE PublicationsGoogle Scholar
  39. Funk JL, Standish RJ, Stock WD, Valladares F (2016) Plant functional traits of dominant native and invasive species in Mediterranean-climate ecosystems. Ecology 97(1):75–83.  https://doi.org/10.1890/15-0974.1 PubMedCrossRefGoogle Scholar
  40. Gelman A (2008) Scaling regression inputs by dividing by two standard deviations. Statist Med 27:2865–2873.  https://doi.org/10.1002/sim.3107 CrossRefGoogle Scholar
  41. Gioria M, Pyšek P (2016) The legacy of plant invasions: changes in the soil seed bank of invaded plant communities. Bioscience 66:40–53.  https://doi.org/10.1093/biosci/biv165 CrossRefGoogle Scholar
  42. Griffin D, Anchukaitis KJ (2014) How unusual is the 2012–2014 California drought? Geophys Res Lett 41:2014GL062433. doi: 10.1002/2014GL062433Google Scholar
  43. Grime JP (1977) Evidence for the existence of three primary strategies in plants and its relevance to ecological and evolutionary theory. Am Nat 111:1169–1194CrossRefGoogle Scholar
  44. Grueber CE, Nakagawa S, Laws RJ, Jamieson IG (2011) Multimodel inference in ecology and evolution: challenges and solutions. J Evol Biol 24:699–711.  https://doi.org/10.1111/j.1420-9101.2010.02210.x PubMedCrossRefGoogle Scholar
  45. Harrison SP, Gornish ES, Copeland S (2015) Climate-driven diversity loss in a grassland community. PNAS 112:8672–8677.  https://doi.org/10.1073/pnas.1502074112 PubMedCrossRefGoogle Scholar
  46. Hobbs RJ, Yates S, Mooney HA (2007) Long-term data reveal complex dynamics in grassland in relation to climate and disturbance. Ecol Mon 77:545–568CrossRefGoogle Scholar
  47. Holling CS (1973) Resilience and stability of ecological systems. Ann Rev Ecol Sys 4:1–23.  https://doi.org/10.1146/annurev.es.04.110173.000245 CrossRefGoogle Scholar
  48. Holmgren M, Stapp P, Dickman CR, Gracia C, Graham S, Gutiérrez JR, Hice C, Jaksic F, Kelt DA, Letnic M, Lima M, López BC, Meserve PL, Milstead WB, Polis GA, Previtali MA, Richter M, Sabaté S, Squeo FA (2006) Extreme climatic events shape arid and semiarid ecosystems. Front Ecol Env 4:87–95.  https://doi.org/10.1890/1540-9295(2006)004[0087:ECESAA]2.0.CO;2 CrossRefGoogle Scholar
  49. Homyak PM, Allison SD, Huxman TE, Goulden ML, Treseder KK (2017) Effects of drought manipulation on soil nitrogen cycling: a meta-analysis. J Geophys Res Biogeosci 2017JG004146. doi: 10.1002/2017JG004146Google Scholar
  50. Hoover DL, Knapp AK, Smith MD (2014) Resistance and resilience of a grassland ecosystem to climate extremes. Ecology 95:2646–2656.  https://doi.org/10.1890/13-2186.1 CrossRefGoogle Scholar
  51. Huenneke LF, Hamburg SP, Koide R, Mooney HA, Vitousek PM (1990) Effects of soil resources on plant invasion and community structure in Californian serpentine grassland. Ecology 71:478–491.  https://doi.org/10.2307/1940302 CrossRefGoogle Scholar
  52. Jiménez MA, Jaksic FM, Armesto JJ, Gaxiola A, Meserve PL, Kelt DA, Gutiérrez JR (2011) Extreme climatic events change the dynamics and invasibility of semi-arid annual plant communities. Ecol Lett 14:1227–1235.  https://doi.org/10.1111/j.1461-0248.2011.01693.x PubMedCrossRefGoogle Scholar
  53. Kimball S, Goulden ML, Suding KN, Parker S (2014) Altered water and nitrogen input shifts succession in a southern California coastal sage community. Ecol App 24:1390–1404.  https://doi.org/10.1890/13-1313.1 CrossRefGoogle Scholar
  54. LaForgia ML, Spasojevic MJ, Case EJ, Latimer AM, Harrison SP (2018) Seed banks of native forbs, but not exotic grasses, increase during extreme drought. Ecology 99:896–903.  https://doi.org/10.1002/ecy.2160 PubMedCrossRefGoogle Scholar
  55. Lauenroth WK, Sala OE (1992) Long-term forage production of North American shortgrass steppe. Ecol App 2:397–403.  https://doi.org/10.2307/1941874 CrossRefGoogle Scholar
  56. Liao C, Peng R, Luo Y, Zhou X, Wu X, Fang C, Chen J, Li B (2008) Altered ecosystem carbon and nitrogen cycles by plant invasion: a meta-analysis. New Phyt 177:706–714.  https://doi.org/10.1111/j.1469-8137.2007.02290.x CrossRefGoogle Scholar
  57. Liu Y, Oduor AMO, Zhang Z, Manea A, Tooth IM, Leishman MR, Xu X, van Kleunen M (2017) Do invasive alien plants benefit more from global environmental change than native plants? Glob Chang Biol 23:3363–3370.  https://doi.org/10.1111/gcb.13579 PubMedCrossRefGoogle Scholar
  58. Loarie SR, Duffy PB, Hamilton H, Asner GP, Field CB, Ackerly DD (2009) The velocity of climate change. Nature 462:1052.  https://doi.org/10.1038/nature08649 PubMedCrossRefGoogle Scholar
  59. Loydi A, Eckstein RL, Otte A, Donath TW (2013) Effects of litter on seedling establishment in natural and semi-natural grasslands: a meta-analysis. J Ecol 101:454–464.  https://doi.org/10.1111/1365-2745.12033 CrossRefGoogle Scholar
  60. Minnich RA, Dezzani RJ (1998) Historical decline of coastal sage scrub in the Riverside-Perris Plain. Western Birds 29:366–391Google Scholar
  61. Moles AT, Gruber MAM, Bonser SP (2008) A new framework for predicting invasive plant species. J Ecol 96:13–17.  https://doi.org/10.1111/j.1365-2745.2007.01332.x CrossRefGoogle Scholar
  62. Oesterheld M, Loreti J, Semmartin M, Sala OE (2001) Inter-annual variation in primary production of a semi-arid grassland related to previous-year production. J Veg Sci 12:137–142.  https://doi.org/10.2307/3236681 CrossRefGoogle Scholar
  63. Ogle K, Barber JJ, Barron-Gafford GA, Bentley LP, Young JM, Huxman TE, Loik ME, Tissue DT (2015) Quantifying ecological memory in plant and ecosystem processes. Ecol Lett 18:221–235.  https://doi.org/10.1111/ele.12399 PubMedCrossRefGoogle Scholar
  64. Pake CE, Venable DL (1996) Seed banks in desert annuals: Implications for persistence and coexistence in variable environments. Ecology 77:1427–1435.  https://doi.org/10.2307/2265540 CrossRefGoogle Scholar
  65. Parker VT, Kelly VR (1989) Seed banks in California chaparral and other Mediterranean climate shrublands. In: Leek MA, Parker VT, Simpson RL (eds) Ecology of soil seedbanks. Academic Press, San Diego, pp 231–255CrossRefGoogle Scholar
  66. Pinheiro J, Bates D, Debroy S, Sarkar D, R Core Development Team (2013) Nlme: linear and nonlinear mixed effects models. R package version 3:1Google Scholar
  67. Potts D et al (2006) Antecedent moisture and seasonal precipitation influence the response of canopy-scale carbon and water exchange to rainfall pulses in a semi-arid grassland. New Phytol 170:849–860.  https://doi.org/10.1111/j.1469-8137.2006.01732.x PubMedCrossRefGoogle Scholar
  68. Powell KI, Chase JM, Knight TM (2011) A synthesis of plant invasion effects on biodiversity across spatial scales. Am J Bot 98:539–548.  https://doi.org/10.3732/ajb.1000402 PubMedCrossRefGoogle Scholar
  69. R Core Development Team (2016) R: a language and environment for statistical computing. R Foundation for Statistical Computing, ViennaGoogle Scholar
  70. Rees M (1994) Delayed germination of seeds: A look at the effects of adult longevity, the timing of reproduction, and population age/stage structure. Am Nat 144:43–64.  https://doi.org/10.1086/285660 CrossRefGoogle Scholar
  71. Reichmann LG, Sala OE, Peters DPC (2013) Precipitation legacies in desert grassland primary production occur through previous-year tiller density. Ecology 94:435–443.  https://doi.org/10.1890/12-1237.1 PubMedCrossRefGoogle Scholar
  72. Reynolds SA, Corbin JD, D’Antonio CM (2001) The effects of litter and temperature on the germination of native and exotic grasses in a coastal California grassland. Madroño 48:230–235Google Scholar
  73. Richardson DM, Pyšek P, Rejmánek M, Barbour MG, Panetta FD, West CJ (2000) Naturalization and invasion of alien plants: concepts and definitions. Div Dist 6:93–107.  https://doi.org/10.1046/j.1472-4642.2000.00083.x CrossRefGoogle Scholar
  74. Robeson SM (2015) Revisiting the recent California drought as an extreme value. Geophys Res Lett 42:2015GL064593. doi: 10.1002/2015GL064593Google Scholar
  75. Rundel PW (2007) Sage Scrub. In: Barbour M, Keeler-Wolf T, Schoenherr AA (eds) Terrestrial vegetation of California, 3rd edn. University of California Press, pp 208–228Google Scholar
  76. Sala OE, Gherardi LA, Reichmann L, Jobbágy E, Peters D (2012) Legacies of precipitation fluctuations on primary production: theory and data synthesis. Philos Trans R Soc Lond B Biol Sci 367:3135–3144.  https://doi.org/10.1098/rstb.2011.0347 PubMedPubMedCentralCrossRefGoogle Scholar
  77. Sandel B, Dangremond EM (2012) Climate change and the invasion of California by grasses. Glob Change Biol 18:277–289.  https://doi.org/10.1111/j.1365-2486.2011.02480.x CrossRefGoogle Scholar
  78. Seabloom EW, Williams JW, Slayback D, Stoms DM, Viers JH, Dobson AP (2006) Human impacts, plant invasion, and imperiled plant species in California. Ecol App 16:1338–1350.  https://doi.org/10.1890/1051-0761(2006)016[1338:HIPIAI]2.0.CO;2 CrossRefGoogle Scholar
  79. Seager R, Ting M, Held I, Kushnir Y, Lu J, Vecchi G, Huang H-P, Harnik N, Leetmaa A, Lau N-C, Li C, Velez J, Naik N (2007) Model projections of an imminent transition to a more arid climate in southwestern North America. Science 316:1181–1184.  https://doi.org/10.1126/science.1139601 PubMedCrossRefPubMedCentralGoogle Scholar
  80. Seebens H, Essl F, Dawson W, Fuentes N, Moser D, Pergl J, Pyšek P, van Kleunen M, Weber E, Winter M, Blasius B (2015) Global trade will accelerate plant invasions in emerging economies under climate change. Glob Change Biol 21:4128–4140.  https://doi.org/10.1111/gcb.13021 CrossRefGoogle Scholar
  81. Soliveres S, Eldridge DJ, Maestre FT, Bowker MA, Tighe M, Escudero A (2011) Microhabitat amelioration and reduced competition among understorey plants as drivers of facilitation across environmental gradients: Towards a unifying framework. Per Plant Ecol Evol Sys 13:247–258.  https://doi.org/10.1016/j.ppees.2011.06.001 CrossRefGoogle Scholar
  82. Sorte CJB, Ibáñez I, Blumenthal DM, Molinari NA, Miller LP, Grosholz ED, Diez JM, D’Antonio CM, Olden JD, Jones SJ, Dukes JS (2013) Poised to prosper? A cross-system comparison of climate change effects on native and non-native species performance. Ecol Lett 16:261–270.  https://doi.org/10.1111/ele.12017 PubMedCrossRefGoogle Scholar
  83. Sousa WP (1984) The role of disturbance in natural communities. Ann Rev Ecol Sys 15:353–391.  https://doi.org/10.1146/annurev.es.15.110184.002033 CrossRefGoogle Scholar
  84. Suding KN, Collins SL, Gough L, Clark C, Cleland EE, Gross KL, Milchunas DG, Pennings S (2005) Functional-and abundance-based mechanisms explain diversity loss due to N fertilization. PNAS 102:4387–4392.  https://doi.org/10.1073/pnas.0408648102 PubMedCrossRefGoogle Scholar
  85. Suttle KB, Thomsen MA, Power ME (2007) Species interactions reverse grassland responses to changing climate. Science 315:640–642.  https://doi.org/10.1126/science.1136401 PubMedCrossRefGoogle Scholar
  86. Talluto MV, Suding KN (2008) Historical change in coastal sage scrub in Southern California, USA in relation to fire frequency and air pollution. Landsc. Ecol. 23:803–815.  https://doi.org/10.1007/s10980-008-9238-3 CrossRefGoogle Scholar
  87. Tilman D (1985) The resource-ratio hypothesis of plant succession. Am Nat 125:827–852CrossRefGoogle Scholar
  88. Underwood EC, Viers JH, Klausmeyer KR, Cox RL, Shaw MR (2009) Threats and biodiversity in the Mediterranean biome. Div Dist 15:188–197.  https://doi.org/10.1111/j.1472-4642.2008.00518.x CrossRefGoogle Scholar
  89. Van Kleunen M, Weber E, Fischer M (2010) A meta-analysis of trait differences between invasive and non-invasive plant species. Ecol Lett 13:235–245.  https://doi.org/10.1111/j.1461-0248.2009.01418.x PubMedCrossRefGoogle Scholar
  90. Vilà M, Espinar JL, Hejda M, Hulme PE, Jarošík V, Maron JL, Pergl J, Schaffner U, Sun Y, Pyšek P (2011) Ecological impacts of invasive alien plants: a meta-analysis of their effects on species, communities and ecosystems. Ecol Lett 14:702–708.  https://doi.org/10.1111/j.1461-0248.2011.01628.x PubMedCrossRefGoogle Scholar
  91. Wainwright CE, Cleland EE (2013) Exotic species display greater germination plasticity and higher germination rates than native species across multiple cues. Biol Inv 15:2253–2264.  https://doi.org/10.1007/s10530-013-0449-4 CrossRefGoogle Scholar
  92. Westman WE (1978) Measuring the inertia and resilience of ecosystems. Bioscience 28:705–710.  https://doi.org/10.2307/1307321 CrossRefGoogle Scholar
  93. Wilsey BJ, Daneshgar PP, Polley HW (2011) Biodiversity, phenology and temporal niche differences between native- and novel exotic-dominated grasslands. Per Plant Ecol Evol Sys 13:265–276.  https://doi.org/10.1016/j.ppees.2011.07.002 CrossRefGoogle Scholar
  94. Wolkovich EM, Lipson DA, Virginia RA, Cottingham KL, Bolger DT (2010) Grass invasion causes rapid increases in ecosystem carbon and nitrogen storage in a semiarid shrubland. Glob Change Biol 16:1351–1365.  https://doi.org/10.1111/j.1365-2486.2009.02001.x CrossRefGoogle Scholar
  95. Yoon J-H, Wang S-YS, Gillies RR, Kravitz B, Hipps L, Rasch PJ (2015) Increasing water cycle extremes in California and in relation to ENSO cycle under global warming. Nature Comm 6:8657.  https://doi.org/10.1038/ncomms9657 CrossRefGoogle Scholar

Copyright information

© Springer Nature B.V. 2019

Authors and Affiliations

  1. 1.Ecology, Behavior & Evolution SectionUniversity of California San DiegoLa JollaUSA
  2. 2.Department of Integrative BiologyUniversity of GuelphGuelphCanada
  3. 3.Department of BiologySan Diego State UniversitySan DiegoUSA

Personalised recommendations