Agricultural and Forest Meteorology 192–193 (2014) 1–8
Contents lists available at ScienceDirect
Agricultural and Forest Meteorology journal homepage: www.elsevier.com/locate/agrformet
Drought events determine performance of Quercus ilex seedlings and increase their susceptibility to Phytophthora cinnamomi Tamara Corcobado, Elena Cubera, Enrique Juárez, Gerardo Moreno, Alejandro Solla ∗ Ingeniería Forestal y del Medio Natural, Universidad de Extremadura, Avenida Virgen del Puerto 2, 10600 Plasencia, Spain
a r t i c l e
i n f o
Article history: Received 7 November 2013 Received in revised form 7 February 2014 Accepted 13 February 2014 Keywords: Climate change Weather extremes Drought Flooding Oak decline Invasive pathogen
a b s t r a c t More frequent weather extremes are expected to occur in the Mediterranean region within the present context of climate change. These extremes could affect forests and plant diseases driven by pathogens. It is hypothesised that simulation of weather extremes during Quercus ilex growth will influence early performance and susceptibility to the invasive oomycete Phytophthora cinnamomi. In 2010, 140 Q. ilex seedlings were subjected to three watering regimes under greenhouse conditions: waterlogging (W), water stress (S) and optimal watering regime for growth (C). During the second vegetative period, conditions were altered to create the following scenarios: WW, WS, SS, SW and CC. After the second vegetative period, plants were artificially infested with P. cinnamomi. Holm oak (Q. ilex) was more sensitive to flooding in the first year of growth than in the second year. The altered scenarios produced plants with a lower fineto-total root ratio than plants in unaltered scenarios. Plants with the highest growth rates maintained their relatively rapid growth and photosynthetic activity under altered scenarios. However, plants with the highest growth rates became the plants with the lowest growth rates when two consecutive years of drought occurred, indicating a trade-off by Q. ilex in growth investment, observed only if the water stress scenario persists. Seedlings were more sensitive to water shortage than to waterlogging, especially if they encountered a dry scenario during the first year. Exposure to drought events increased seedling mortality rates after P. cinnamomi infection. Waterlogging combined with subsequent water deprivation was the worst scenario when soil was infested with P. cinnamomi, causing 100% mortality of plants. © 2014 Elsevier B.V. All rights reserved.
1. Introduction The impacts of recent global climate change have been well illustrated by shifts in plant phenology, morphology, genetic frequencies and abundance, with a range of wild biota moving poleward or upward (Richardson et al., 2013). Direct harm to forests caused by extreme climate events has been reported (Lloret et al., 2004; Lindner et al., 2010). Considering the modern climate change scenarios predicted for the next century, new environments will be expected, with the subsequent difficulty of anticipating how species will respond to these changes (Visser, 2008). Plant diseases driven by pathogens are expected to exacerbate the negative effects of climate change on forests (Santini et al., 2013). Pathogen distribution and development are especially limited by temperatures for overwintering and oversummering, but with higher winter temperatures under global warming, an expansion of range is expected and
∗ Corresponding author. Tel.: +34 927257000/+34 654899297; fax: +34 927425209. E-mail address:
[email protected] (A. Solla). http://dx.doi.org/10.1016/j.agrformet.2014.02.007 0168-1923/© 2014 Elsevier B.V. All rights reserved.
latency stages may be shorter. In some cases, higher occurrence of extreme events would allow pathogens to increase their number of cycles per year. Literature reviews of the impacts of climate change on plant diseases appear to be based on models rather than on research studies using real data (Pautasso et al., 2012). The Mediterranean regions are particularly vulnerable to global change, according to climate model simulations that consistently project pronounced drying and warming (Giorgi and Lionello, 2008) and predict more frequent extreme events of drought and waterlogging (Lindner et al., 2010). In this region, the invasive pathogen Phytophthora cinnamomi is the main biotic factor involved in oak decline (Sánchez et al., 2005; Brasier, 1996; Solla et al., 2009; Corcobado et al., 2013a, 2013b), together with other recently described soilborne Phytophthora species (Corcobado et al., 2010; Pérez-Sierra et al., 2013). The consequences of Phytophthora infections on plants are similar, but not identical, to those resulting from severe water stress (Manter et al., 2007; Oßwald et al., 2014). Because this pathogen is highly dependent on soil water content and mild temperatures (Corcobado et al., 2013a; McConnell and Balci, 2014), a combination of warming and weather extremes is likely to enhance P. cinnamomi dispersal and impact.
2
T. Corcobado et al. / Agricultural and Forest Meteorology 192–193 (2014) 1–8
Tree resistance to drought and other disturbances can be ˜ enhanced during the seedling stage (Banón et al., 2006; VillarSalvador et al., 2012). Plant performance is also affected in response to hardening practices, showing increased root and shoot growth or a lower shoot-to-root ratio (van den Driessche, 1992). The increased duration, intensity and frequency of drought events expected in Mediterranean areas could contribute to the hardening of trees, as a recent study has shown in relation to frost (Thorsen and Höglind, 2010). Varied plant responses would be associated with the later dehardening period and subsequent improved plant performance involving, for example, better control of water loss or increased root regeneration (Grossnickle, 2012). Ameliorated performances after hardening could result in plant tolerance to biotic factors, especially those causing symptoms and physiological disturbances similar to drought. After first infection, plants react with a variety of strategies to subsequent infections. These strategies may also be activated or induced by previous biotic and abiotic stressors (Eyles et al., 2010; Vivas et al., 2013). Water deficit and heavy rains are common phenomena in Mediterranean areas, but their role as inducers has seldom been reported (Solla and Gil, 2002; Sun et al., 2010). Within the present context of global climate change, more frequent drought and flooding phenomena make it necessary to study these scenarios as inducers of resistance or susceptibility. Quercus ilex is a circun-mediterranean tree that extends 6000 km longitudinally from Portugal to Syria, and 1500 km latitudinally form Maghreb to France. This evergreen tree covers more than 4 million ha under a wide range of edaphoclimatic conditions, especially in the western Mediterranean countries. In Spain and Morocco Q. ilex extends over ca. 30% of the natural forest cover and is extensively used in artificial reforestation programmes (González-Rodríguez et al., 2011). Artificial methods of Q. ilex regeneration include direct seeding of acorns and nursery production of seedlings. Direct-seeded plants grow to a lesser extent, are more vulnerable to predation, and survive less than 1-year-old planted seedlings (González-Rodríguez et al., 2011), and for these reasons the use of seedlings is recommended. Plantations and natural forests of Q. ilex have been severely damaged by P. cinnamomi infections, being Q. ilex the oak species more susceptible to P. cinnamomi (Robin et al., 2001). However, the susceptibility of Q. ilex to climate extremes followed by P. cinnamomi infections is unknown. It is hypothesised that simulation of weather extremes during Q. ilex seedling growth will influence early seedling performance and susceptibility to P. cinnamomi. In this study, Q. ilex seedlings were exposed to changing watering conditions that simulated weather extremes and after two vegetative periods, plants were inoculated with P. cinnamomi. Plant growth, mortality and physiological functioning were assessed to answer the following questions (i) are Q. ilex seedlings able to adapt to prolonged drought and/or waterlogging events through physiological and growth adjustments? (ii) which combination of drought and waterlogging events is more closely associated with higher mortality rates? and (iii) will these extreme events affect seedling responses to subsequent P. cinnamomi infections?
2. Materials and methods During the first vegetative period, Q. ilex seedlings were exposed to conditions of optimal watering regime for growth (C), prolonged waterlogging (W) and prolonged water stress (S). W and S simulated extreme precipitation and extreme Mediterranean summer drought events, respectively. During the second vegetative period, these scenarios were maintained (CC, WW and SS) and also crossed over (WS and SW), as seen in Fig. 1. At the end of the second vegetative period, all seedlings were artificially inoculated with P. cinnamomi, simulating introduction of an invasive pathogen (Fig. 1).
2.1. Plant material The plant material came from a Q. ilex savannah-like forest (Malpartida de Plasencia, south-west Spain, 39◦ 07 N, 7◦ 29 W; 314 m asl). In November 2009, acorns were collected from a single Q. ilex tree and stored in a cold chamber at 4 ◦ C for three months. In February 2010, the acorns were germinated on trays with a mixture of peat and vermiculite (3:1) and watered close to field capacity. A week later, acorns with emerging radicles were individually planted in 140 cylindrical PVC pots (approximately 16 l volume; 1.50 m high, 11.5 cm inner diameter; Fig. 2a ) and packed with sand and peat (1:1, pH 5.5). Earlier research by Cubera et al., 2009 showed that this pot size would provide seedlings with unrestricted root growth during the experiment. Plastic mesh was placed at the bottom of all pots to prevent substrate movement and facilitate water draining and root air pruning. To avoid water loss during waterlogging, pots were sealed with long plastic bags measuring 150 cm. The outer walls of the pots were painted white to prevent them from being warmed by the sun. Pots were kept in natural daylight under greenhouse shade that reduced solar radiation by 50%, at the Plasencia Faculty of Forestry (University of Extremadura), Spain (40◦ 02 N, 6◦ 05 W; 374 m asl) (Fig. 2a). 2.2. Watering treatments Every 3–4 days, all pots were initially hand watered to field capacity until the plants were well established (Fig. 2b). In May 2010, when seedlings had emerged after three months of growth, a drip-irrigation system was installed and pots were divided into C, W and S groups according to the three watering treatments (Fig. 1). Seedlings subjected to the various watering regime scenarios were distributed at random. For 2.5 months 20 pots were watered every two days with 200 ml per pot (C regime), 60 pots were kept waterlogged (W), and 60 pots were not watered except for one day in mid July, with 2000 ml per pot (S). After 2.5 months all plants were watered as C. In May 2011, some plants subjected to waterlogging and water stress treatments were changed over to water stress and waterlogging treatments, respectively, and for 2.5 months five treatments were established: CC, WW, WS, SS and SW (n ≥ 17) (Fig. 1). Soil moisture was checked in all pots at the end of both 2010 and 2011 watering treatments (mid-August) at depths of 0.50 and 100 cm using a Delta-Theta ML2X probe, through bores made in the wall of the PVC pots. Measurements confirmed differences between watering regimes (Fig. 3). Higher soil moisture values were observed under seedlings subjected to W and SW treatments (2010 and 2011, respectively) than under seedlings subjected to other treatments (F = 192.3, p > 0.001 and F = 135.7, p > 0.001; Fig. 3). 2.3. P. cinnamomi inoculum preparation and soil infestation The P. cinnamomi strain used in the experiment was isolated from Badajoz during surveying by Corcobado et al., 2013b. Inoculum was prepared following the procedure of Jung et al., 1996, mixing and twice autoclaving 500 cm3 fine vermiculite, 40 cm3 whole oat grains and 350 ml multivitamin juice broth (200 ml l−1 juice, 800 ml l−1 demineralised water amended with 3 g l−1 CaCO3 ) in 1l Erlenmeyer flasks. Individual pieces of P. cinnamomi plugs were then added to the Erlenmeyer flasks containing the medium and kept in an incubator at 20 ◦ C for five weeks. Control flasks containing the same medium without P. cinnamomi plugs were also incubated. The soil infestation process was conducted in October 2011, when Q. ilex plants were about two years old. The inoculum was rinsed with demineralised water to remove excess nutrients. Seedlings were then carefully transplanted into larger PVC tubes (15.5 cm in diameter) containing 25 l substrate (sand and peat 1:1,
T. Corcobado et al. / Agricultural and Forest Meteorology 192–193 (2014) 1–8
2009 Acorn sampling
2010 Germination Regular watering (Control) n=20
2011
3
2012 Inoculations
Survival assessment
Regular watering (CC) n=20
Regular watering n=16
Waterlogging (WW) n=17
Regular watering n=12
Water stress (WS) n=18
Regular watering n=12
Water stress (SS) n=20
Regular watering n=12
Waterlogging (SW) n=20
Regular watering n=9
Waterlogging (W) n=60
Water stress (S) n=60
Fig. 1. Experimental design showing five watering scenarios in which Quercus ilex seedlings grew during two vegetative periods. Soil infestation with Phytophthora cinnamomi (inoculations) was performed after the second vegetative period. Plant replicates are shown.
pH 5.5) mixed at 2% with the inoculum or with the non-infested medium (control tubes). Tubes were placed in a very large container (2000 l) to allow complete flooding (Fig. 2d). After transplanting, seedlings received normal drainage and were then flooded the following day. Every three weeks until summer 2012, seedlings were waterlogged for 72 h to stimulate P. cinnamomi sporulation and zoospore release (Thomas Jung, personal communication). In October 2012, P. cinnamomi was successfully reisolated from root samples collected from the artificially infested soil. Rootlets were cut into 1 cm segments, surface sterilised (2 min in 1% aqueous sodium hypochlorite), rinsed with sterile water, blotted dry and plated onto the selective medium. Approximately 15 plates per tube were incubated in the dark at 24 ◦ C and after 2–3 days selected isolates were transferred to a carrot agar (CA) medium. Colonies were identified by microscopic observations of distinctive structures such as clustered hyphal swellings, chlamydospores and sporangia.
2.4. Plant measurements In February 2010, before germination, acorns were individually weighed. Plant survival was recorded monthly, but after the second vegetative period, mortality was monitored weekly. Seedlings were recorded as dead when they had lost all their leaves, had no green or flexible leaves, or exhibited loss of stem flexibility (Fig. 2e, Valladares and Sánchez-Gómez, 2006). Plant height and number of leaves were assessed at the end of watering treatments in both 2010 and 2011. In July 2010 and 2011, stomatal conductance (gs , mol·H2 O m−2 s−1 ) and net leaf photosynthesis (A, mol CO2 m−2 s−1 ) were determined on approximately eight plants per treatment using a portable differential infrared gas
analyser (IRGA) (LCi, ADC Bio Scientific Ltd., UK) connected to a broadleaf chamber. Gas exchange was measured from 9.30 to 11.00 h in one current-year leaf per plant at saturating light (1500 mmol m−2 s−1 ) with a red/blue light emitting diode (LC pro, ADC). In August 2011, a subsample of plants was used for above- and below-ground assessment. Four plants per treatment (20 plants in total) were removed from the tubes. Their shoots were cut at the cotyledon insertion point and shoot weight and length, number of leaves and leaf weight were assessed. For root system assessment, tubes were opened with a radial saw (Fig. 2c) and the cylinders of soil containing the plants were extracted and cut into six sections. Fine and coarse roots were collected with a sieve and tweezers and then weighed.
2.5. Data analysis To assess the influence of watering treatments on parameters of physiology (gs and A), growth (stem height, number of leaves and annual aboveground growth) and root system (root biomass, fine-to-total root ratio, root-to-shoot ratio and fine root-to-leaf ratio), one-way ANOVAs were performed using each parameter as the dependent variable and 2010 or 2011 watering treatments as the fixed factor. The relation between these parameters was analysed using Pearson correlations. To analyse how the watering regimes affected the relations between plant growth rate parameters, homogeneity of slopes tests were performed. Normality and homoscedasticity of the data were checked using Kolmogorov-Smirnov and Bartlett’s tests. The experiment ended at the beginning of autumn 2012, 48 weeks after plants had been inoculated with P. cinnamomi. To
4
T. Corcobado et al. / Agricultural and Forest Meteorology 192–193 (2014) 1–8
Fig. 2. Changing scenario experiment in which (a) 140 white painted PVC tubes 1.5 m high were used, (b) Quercus ilex seedlings were grown inside, and (c) some were opened for root system analysis. A large container (d) allowed periodic waterlogging of tubes and seedlings to stimulate cycles of Phytophthora cinnamomi sporulation and zoospore release until (e) plants were recorded dead.
analyse time-to-death of plants (survival time) and determine survival probabilities of seedlings subjected to the various watering regimes and seedlings inoculated with P. cinnamomi, the Kaplan–Meier estimate was used. This survival analysis technique is a non-parametric procedure in which seedlings that were alive at the end of the experiment were considered censored because their time-to-death was unknown (Solla et al., 2011). Survival analysis is a method that enables the analysis of all the survival data, i.e. seedlings that were infected and died, plus seedlings that were infected but did not die by the end of the study (Esker et al., 2006; Vivas et al., 2012). This method offers several advantages over ANOVA approach methods since it handles repeated measurements over time on the same sampling units, including censored observations, and accounts for failure times (i.e., death) that are not normally distributed. For the comparison of survival times of seedlings between CC, WW, WS, SS and SW treatments, a non-parametric test was used and the p value provided. This multiple-sample test consisted of an extension (or generalisation) of Gehan’s generalized Wilcoxon test, Peto and Petos’s generalised Wilcoxon test, and the log-rank test. Additionally, mean survival
times (±SE) of seedlings from each treatment were determined. To test the influence of acorn size, plant size and plant growth rate on seedling survival, generalised linear models were performed using mortality (parameterised as 0 or 1 if the seedling was alive or dead, respectively) after the watering treatments or after P. cinnamomi infestations as the dependent variable, 2011 watering treatments as the fixed factor and stem height, stem growth or acorn weight as covariates. The Likelihood type I test was used within these models. All analyses were conducted with STATISTICA v.10 software.
3. Results In 2010, seedlings subjected to W and S treatments showed significantly lower rates of CO2 assimilation (A) than seedlings under C treatment (3.1 ± 0.4, 3.4 ± 0.4 and 4.7 ± 0.4 mol CO2 m−2 s− 1, respectively; F = 4.77, p = 0.014). Stomata conductance (gs ) was slightly lower in seedlings under W and S treatments than seedlings under C treatment (F = 2.70, p = 0.079). Mean stem height was 23.7 and 26.3% lower in
T. Corcobado et al. / Agricultural and Forest Meteorology 192–193 (2014) 1–8 10
Soil moisture (%vol) 0
10
20
30
0
Depth (cm)
25 C W
50
S 75 100
CC
9
WW
Aboveground growth during 2011 (g)
(a)
8
WS
7
SS
(CC) y = 3.43±0.77 x + 1.15±0.66 R² = 0.53
SW
6 5
(WW) y = 2.53±0.87 x + 1.17±0.55 R² = 0.38
4 3
(WS) y = -0.01±0.53 x + 1.59±0.34 R² = 0.00
2 1
(SW) y = -1.10±1.23 x + 1.65±0.49 R² = 0.07 (SS) y = -2.06±0.61 x + 1.75±0.16 R² = 0.45
0
Soil moisture (%vol)
(b) 0
10
20
0
30
Depth (cm)
0 25
CC WW
50
WS SS
75
5
SW
100 Fig. 3. Mean soil moisture values (±SE) in (a) 2010 and (b) 2011 of pots containing Quercus ilex seedlings. Plants were subjected to an optimal watering regime for growth (CC) and combined waterlogging (W) and water stress (S) scenarios.
seedlings under W and S treatments than seedlings under C treatment (F = 9.21, p < 0.001). Similarly, the number of leaves of seedlings subjected to W and S treatments was 14.5 and 19.0% lower than seedlings under C treatment (F = 3.56, p = 0.031). In 2011, the highest and lowest A values were observed in plants subjected to CC and SW, respectively, and for gs , the highest and lowest values were also observed in plants subjected to CC and SW (F = 8.34, p < 0.001 and F = 7.25, p < 0.001, respectively; Table 1). Photosynthetic activity was positively related to plant growth: in 2011 A was positively correlated to the number of leaves per plant (rpearson = 0.48; p = 0.002; n = 40) and aboveground growth rate (rpearson = 0.38; p = 0.002; n = 40). In 2011, height was lower for all treated plants than for CC plants (F = 4.60, p = 0.002; Table 1), especially those first treated with S. Plants under SW treatment had approximately 45% fewer leaves than plants under WW and CC treatments (F = 3.41, p = 0.005). Considering stem and leaf biomass together, aboveground growth in 2011 was highest in CC plants (Table 1). For CC plants, aboveground growth in 2011 was positively related to aboveground growth in 2010 (rpearson = 0.79; p < 0.0001; n = 20), but this relation varied significantly depending on the watering scenarios applied in 2011 (significant watering treatments × above ground growth interaction; F = 3.98; p = 0.006) and was negative for SS plants (rpearson = −0.60; p < 0.0149; n = 16; Fig. 4). In 2011, root system biomass did not change significantly between treatments (F = 1.60; p = 0.196; Table 1), but the ratio of fine root biomass to total root biomass was lower for plants under changing scenarios than for plants under a constant scenario (F = 3.14, p = 0.025; Table 1). In 2010, plant mortality was higher under W scenarios (15%) than under S and C scenarios (1.7% and 0.0%, respectively; Table 2). In 2011, scenarios which had included the S treatment in 2010 (SS and SW; Table 2) had the highest mortality rates. The drought
0,5
1
1,5
2
Abovegroung growth during 2010 (g)
Fig. 4. Relations between aboveground growth (stem + leaf biomass) of Quercus ilex plants in 2010 and 2011 after being subjected to five different watering regimes (Fig. 1). Homogeneity-of-slopes tests were highly significant for watering treatment × aboveground growth interaction (F = 3.98; p = 0.006; d.f. = 4 and 70).
event in 2010 resulted in more delayed and higher plant mortality than the optimal watering regime for growth and the waterlogging scenarios (Fig. 5). Plant mortality was independent of acorn size (2 = 0.10; p = 0.755), regardless of the scenario (2 = 0.21; p = 0.976 for the watering treatments × acorn weight interaction). After P. cinnamomi inoculations, maximum mortality rates were observed in plants subjected to water stress treatment in 2011 (WS and SS) and lower mortality rates were observed in plants subjected to WW and SW treatments (Table 2; Fig. 6). Mean survival times (±SE) of seedlings treated as CC, WW, WS, SS and SW were 31.1 ± 7.9, 38.8 ± 2.7, 25.2 ± 4.8, 30.8 ± 4.3 and 15.3 ± 10.0 weeks, respectively. Waterlogging combined with subsequent water deprivation and P. cinnamomi infestation was the worst scenario for Q. ilex survival, resulting in 100% mortality of plants. Mortality was independent of acorn size (2 = 0.45; p = 0.502), regardless of the scenario (2 = 0.39; p = 0.942 for the watering treatments × acorn weight interaction).
Fig. 5. Plot of survival probabilities using the Kaplan–Meier estimate, which showed differences in plant mortality rates after 2011 watering treatments (p = 0.037). X axis, months from the beginning of watering treatments (May 2010); Y axis, proportion of surviving plants.
6
T. Corcobado et al. / Agricultural and Forest Meteorology 192–193 (2014) 1–8
Table 1 Mean values (±standard error) and significances of parameters related to physiological activity, aboveground growth and root system of Quercus ilex seedlings growing under optimal watering regime for growth (CC) and combined waterlogging (W) and water stress (S) scenarios (Fig. 1).
Physiology A, mol CO2 m−2 s−1 gs, mmol·H2 O m−2 s−1 Growth Stem height (cm) Above ground growth (g) Root system Root biomass (g) Fine-to-total root ratio Root-to-shoot ratio Fine root-to-leaf ratio
CC
WW
WS
SS
SW
F
p Value
6.4 ± 1.8c 0.11 ± 0.02c
2.9 ± 0.7ab 0.05 ± 0.01a
2.7 ± 0.5ab 0.05 ± 0.01ab
3.9 ± 0.4b 0.08 ± 0.01b
1.6 ± 0.4a 0.03 ± 0.01a
7.25 8.34
