Transgenerational effect alters the interspecific competition between two dominant species in a temperate steppe

Abstract One of the key aims of global change studies is to predict more accurately how plant community composition responds to future environmental changes. Although interspecific relationship is one of the most important forces structuring plant communities, it remains a challenge to integrate long‐term consequences at the plant community level. As an increasing number of studies have shown that maternal environment affects offspring phenotypic plasticity as a response to global environment change through transgenerational effects, we speculated that the transgenerational effect would influence offspring competitive relationships. We conducted a 10‐year field experiment and a greenhouse experiment in a temperate grassland in an Inner Mongolian grassland to examine the effects of maternal and immediate nitrogen addition (N) and increased precipitation (Pr) on offspring growth and the interspecific relationship between the two dominant species, Stipa krylovii and Artemisia frigida. According to our results, Stipa kryloii suppressed A. frigida growth and population development when they grew in mixture, although immediate N and Pr stimulated S. kryloii and A. frigida growth simultaneously. Maternal N and Pr declined S. krylovii dominance and decreased A. frigida competitive suppression to some extent. The transgenerational effect should further facilitate the coexistence of the two species under scenarios of increased nitrogen input and precipitation. If we predicted these species' interspecific relationships based only on immediate environmental effects, we would overestimate S. krylovii's competitive advantage and population development, and underestimate competitive outcome and population development of A. frigida. In conclusion, our results demonstrated that the transgenerational effect of maternal environment on offspring interspecific competition must be considered when evaluating population dynamics and community composition under the global change scenario.


| INTRODUC TI ON
The global environment is undergoing rapid change, owing to a variety of natural and anthropogenic events (MEA, 2005). Concurrent with climate warming, precipitation is projected to increase in midlatitude terrestrial ecosystems (Donat et al., 2016;IPCC, 2013;Knapp et al., 2015). As fertilization or atmospheric deposition, nitrogen input into terrestrial ecosystems has been increased from 34 to 100 Tg N/year, and it will reach 200 Tg N/year by 2050 (Galloway et al., 2008). One of the central goals of global change studies is to accurately predict how plant diversity and community composition respond to future environmental changes (Bellard et al., 2012;Dawson et al., 2011;Gilman et al., 2010;Moritz & Agudo, 2013;Zhang et al., 2019). It is generally agreed that interspecific competition is among the most important forces structuring plant communities (Aerts, 1999;Niu & Wan, 2008) and is commonly environment-dependent (Gilman et al., 2010). Studies on transgenerational effects have shown that maternal environment affects offspring performance (Franks et al., 2007;Galloway, 2001;Huxman et al., 2001;Miska & Ferguson-Smith, 2016). However, little is known about how the maternal environment influences interspecific competition through the transgenerational effect. Because offspring may exhibit different plasticities as a response to environmental changes from that of the maternal plant through transgenerational effect (Lau et al., 2008;Li et al., 2011), the interspecific relationship of offspring plants in response to environmental change may also be different from that of the maternal plant. Therefore, to accurately evaluate future response of community composition to environmental change, it is important to ascertain how maternal nitrogen addition or/and increased precipitation affect interspecific competition among maternal plants or among offspring plants.
The transgenerational effect could alter interspecific competition by altering offspring performance and plasticity (Miska & Ferguson-Smith, 2016;Rasanen & Kruuk, 2007). As sessile organisms with limited escape ability (Holeski et al., 2012), plants have to adapt or they become extinct as a response to environmental changes (Yin et al., 2019). According to studies on transgenerational effects, plants not only have enormous potential for phenotypic plasticity to immediate environmental change (Neytcheva & Aarssen, 2008), but they also transfer plasticity to their offspring (Franks et al., 2007;Galloway, 2001;Huxman et al., 2001;Li et al., 2011Li et al., , 2017Li et al., , 2019. The transgenerational effect of maternal environment and plasticity effect of the immediate environment both influence offspring performance (Auge et al., 2017), although the direction and magnitude of the two effects may be different (Lau et al., 2008;Li et al., 2011). Consequently, competitive ability, which is calculated for each trait, may be simultaneously affected by maternal environment and immediate environment. It is essential to quantify competitive ability or outcome not only through the plasticity effect but also through the transgenerational effect because being a good competitor or surviving under competition may occur via pluralistic approaches in the parental generation or in filial generations (Aarssen & Keogh, 2002;Aschehoug et al., 2016;Hart et al., 2018;MacDougall & Turkington, 2004;Wang et al., 2010). Therefore, competitive outcome in response to immediate environment only, which has been tested in many previous studies, is insufficient to represent long-term competitive outcomes and consequent community composition (Freckleton et al., 2009;Hart et al., 2018;Laughlin, 2014;Tylianakis et al., 2008;Yin et al., 2019).
Theoretical (Ezard et al., 2014;Galloway & Etterson, 2007) and experimental (Burgess & Marshall, 2011) studies have shown that the transgenerational effect on offspring is adaptive when the environment is predictable (Uller, 2008). In addition, the environment could be anticipated because environmental change is correlated under the global change scenario. If offspring traits can be affected by transgenerational effects, considering that the transgenerational effect of each species in a community is adaptive, how does interspecific competition change and how does community composition change as a consequence? Relatively, little is known about how the maternal environment influences interspecific competition during the adaptation of each species to a changing environment.
Therefore, quantifying the effect of the maternal environmental effect on interspecific competition is crucial for accurately evaluating the future response of community composition to global changes (Auge et al., 2017).
As an important biome, grasslands occupy 36% of the terrestrial land cover (Sala, 2001). Quantifying the magnitude of the responses of grass community composition to nitrogen and rainfall alterations is crucial for accurately evaluating the global response of plant productivity and diversity to future climate scenarios . To experimentally elucidate whether the maternal environment influences offspring interspecific competition through transgenerational effect, a field and a greenhouse experiment were conducted in a typical steppe in Inner Mongolian grassland. The field experiment was part of a multifactor field experiment initiated in April 2005 Yang et al., 2012). Increased precipitation and N addition have consistently been documented to profoundly impact community composition and biodiversity in this steppe ecosystem (Xu et al., 2018;Yang et al., 2011Yang et al., , 2012. Moreover, pioneer studies at the site have demonstrated that maternal-increased precipitation and nitrogen environments play important roles in offspring performance through transgenerational effects (Li et al., , 2017. Despite the growing evidence on rapid adaptive evolution as a response to environmental changes in the plant population, the consequences of such rapid evolution on plant diversity and community composition remain to be explored (Lavergne et al., 2010). The competitive relationship of dominant species is important for understanding community composition and biodiversity in ecosystems (Fay et al., 2003;Zhang et al., 2017). Thus, Stipa krylovii and Artemisia frigida, two dominant species that coexist in the study ecosystem (Liu et al., 2016), were selected in our study. The specific objectives of our study were to reveal (a) how immediate nitrogen addition, increased precipitation, and their interaction affect interspecific competition between the two dominant species through phenotypic plasticity; and (b) whether and how maternal nitrogen addition, increased precipitation, and their interaction affect interspecific competition between the two dominant species through the transgenerational effect.

| Plant material
Stipa krylovii and A. frigida, two perennial species in the temperate steppe, were selected for this study. Stipa krylovii, a tall bunchgrass, is widespread in arid and semiarid grasslands in Inner Mongolia, China.
This species flowers, from the end of July to August, and its seeds are dispersed in September. At the peak of the growing season, it can grow to approximately 50 cm. Artemisia frigida, a semishrub, occupies approximately 38% of the total foliar biomass in the steppe ecosystem. This species flowers at the end of August and its seeds are dispersed in October. During the peak of the growing season, its vegetative tiller is usually less than 10 cm and its reproductive tiller can reach 30 cm in height.

| Field experimental design and seed collection
The permanent site of the Duolun Global Change Multifactor Experiment (GCME) was established in 2005 in northern China Yang et al., 2011Yang et al., , 2012. GCME employed a design with N addition manipulated at the plot level and precipitation manipulated at the subplot level. Four pairs of 44 × 28 m plots were established; one plot of each pair was randomly assigned as control (C) and the other as "N addition" (N) treatment. Nitrogen at 10 g N/m 2 (urea in 2005 and NH 4 NO 3 in 2006-2015) was applied in the N addition treatments in mid-July every year. In each control or N addition plot, two 10 × 15 m subplots were set up, of which one was watered in summer and the other was not. From July to August every year, 15 mm of water was added once a week, leading to an annual addition of 120 mm precipitation (approximately 30% of the mean annual precipitation in the study site) in a water addition subplot.
From September to October 2015, seeds of S. krylovii and A. frigida were collected from GCME. Before seed diffusion, seeds were collected throughout the whole plot or subplot as evenly as possible. For one species, seeds collected from the plots or subplots of each treatment in the field were mixed together. Treatments in GCME were considered to be the maternal environment in our analysis and were named as follows: maternal control environment

| Greenhouse experiment
A greenhouse experiment was conducted adjacent to the field site.
To be consistent with the environmental conditions of GCME, the greenhouse wall and ceiling were opened on sunny days and closed the greenhouse on rainy days. During the entire experimental period, the temperature of the greenhouse was approximately 22.7°C in the daytime and approximately 12.3°C at night.
After being air-dried, seeds collected in GCME (treatments: were transplanted into pots. All pots received the same treatment. Each pot was watered with 100 ml to maintain growth every day and received about 600 mm water during the entire growth season (from May to August). These values were higher than the MAP in Duolun County because the sandy soil in small pots and higher temperature in the greenhouse required more water .
That is, all the pots used to this point of the experiment were under the same immediate environment; thus, the seedlings from the four maternal environments received the same immediate environment. Furthermore, each pot had two seedlings: two S. krylovii seedlings (monoculture of S. krylovii), two A. frigida seedlings (monoculture of A. frigida), or one S. krylovii seedling and one A. frigida seedling (mixed cultures of the two species). Seedlings in a pot were from the same maternal environment. Six replications were set for each treatment.
Second, in order to observe the effect of the immediate environment on interspecific competition, seedlings from maternal control treatment (M-CK) in GCME were transplanted into pots in the greenhouse. The seedlings were randomly assigned to four immediate environments crossed with three competition treatments (monoculture of S. krylovii, monoculture of A. frigida, or mixed cultures of the two species) within a block. The four immediate environments that simulated maternal environments in GCME were applied as control, nitrogen addition, increased precipitation, and N addition and increased precipitation treatments to seedlings in this part of the experiment. To differentiate these treatments from those simulating the maternal environments, we named these four immediate environments as I-CK, I-N, I-Pr, and I-NPr, respectively. Each pot was watered with 100 ml of water per day to maintain plant growth.
Beginning in late May, water was replaced with 100 ml of 0.5 g/L NH 4 NO 3 solution, which was added to each N addition treatment alone (I-N) and to the N plus water addition pot (I-NPr); this was performed 10 times during a 10-day interval until late August for one year. The precipitation treatment was measured in August with an additional 100 ml of water every day in each I-Pr or I-NPr pot. A total of 200 ml of water per day was added to the precipitation-increased pots in order to make a substantial difference in water availability compared with the ambient watering treatments (100 ml water per day) because of the sandy soil and higher temperature in the greenhouse . That is, the seedlings from the M-CK treatment received four immediate environments in this part. There were six replications for each treatment.
We harvested plants in late August 2017. At harvest, shoots of one individual were clipped at the soil surface and placed in a paper bag. To identify the shoots, we labeled the roots before clipping the two individuals in monoculture pots. The roots and soil from each pot were placed in plastic mesh bags. The soil was washed from the roots with flowing water. Thereafter, the roots of one individual were placed in a paper bag. Because of the low plant density, sandy soil, and root system morphology differences between the two species, it was feasible to assign roots to individuals and ensure root extraction efficiency (Cahill, 2002;Wang et al., 2010). The dry mass was oven-dried at 65°C for 48 hr to a constant weight.

| Statistical analysis
We used competitive response (CR) to determine competitive ability in our study as CR is highly labile and contingent upon the environment (Goldberg & Fleetwood, 1987;Wang et al., 2010). For each species, CR was measured using the species' biomass when grown with other species (in mixture) divided by its biomass grown with the same species (in monoculture). For each species, seedlings with approximate shoot size, root size, and leaf number before transplanting were chosen. We calculated the CR of each species using the biomass of the species in mixture divided by the biomass in monoculture randomly. As there were six replications for each treatment, there were also six replications for CR. When the CR of S. krylovii was calculated, S. krylovii was considered the target species and A. frigida was the accompanying species. The inverse was also performed. A high CR value indicates a strong ability of the target species to mitigate the cost of competition with the accompanying species (Wang et al., 2010).
Three-way ANOVAs were used to examine the main and interactive effects of N addition, increased precipitation in maternal or in immediate environments, competition on offspring biomass, and biomass allocation of the two species. Two-way ANOVAs were used TA B L E 1 Effects of immediate nitrogen addition (N), immediate increased precipitation (Pr), competition (C), and their interactions on biomass, aboveground biomass, belowground biomass, and S/R of the two species based on three-way ANOVAs to examine the main and interactive effects of N addition and increased precipitation in the maternal or immediate environment on the CR of the two species. All statistical analyses were conducted using the SAS software (SAS Institute Inc.).

| Immediate environment affected the interspecific relationship of the two species
For S. kryroii, competition significantly affected biomass accumulation and allocation (all p < .05), while immediate nitrogen and water addition only significantly affected some aspects of biomass, especially aboveground biomass. Immediate nitrogen addition significantly increased total biomass by 71.82% (F = 13.21, p < .001; Table 1, Figure 1). There were significant interactive effects between immediate nitrogen addition and water addition on aboveground biomass and S/R (F = 53.78, p < .001; F = 11.42, p < .05;

| Maternal environment affected the species' interspecific relationship
For S. kryroii, competition and maternal nitrogen addition had no significant effect on biomass and biomass allocation (all p > .05,

| Immediate environment affected the species' interspecific relationship
Our results showed that environmental change affected the interspecific competition between the two dominant species. Compared  Our results showed that the two dominant species exhibited different strategies to promote competitive ability. In stressful environments, species exhibit phenotypic plasticity in the life history to promote competitive ability, such as alteration of one or a combination of these three fundamental traits: growth, survival, and fecundity (Bolker & Pacala, 1999;Drenovsky et al., 2008;Wang et al., 2010). Furthermore, because of their different intrinsic morphological attributes, plants can choose different strategies to maintain growth, survival, and fecundity (Liu et al., 2016). They may allocate new biomass to leaves and/or roots to increase the plant's capacity to acquire resources (Chapin, 1991;Hermans et al., 2006), or to organs such as stolons, rhizomes, and seeds to increase the ability to exploit new patches (Drenovsky et al., 2008). According to our results, as a gramineous bunch grass, S. krylovii significantly stimulated aboveground biomass with nitrogen and water supply in order to intercept light to maintain growth and survival, suppressing A. frigida growth in mixture (Niu et al., 2009). For A. frigida, although the aboveground and belowground biomass both increased under N addition and increased precipitation, the biomass, especially biomass allocation to shoots, significantly decreased when grown with S. krylovii. As a shorter and stoloniferous clonal forb, A. frigida increased belowground biomass allocation to capture more resources in soil to maintain growth and survival. Our results were in line with the fact that an inferior competitor could reduce competitive suppression by accessing resources not available to a superior competitor (Carmona et al., 2019). Moreover, although the aboveground biomass of A. frigida decreased by 60.09% in the mixture (Figure 1), the largest length of stolon increased by 13.93% in the mixture (Li Y, unpublished data). This means that A. frigida in the mixture might allocate more biomass to the stolon for exploring new patches. Based on the size dependence of individual-level resource use and architecture (Allen et al., 2008), the allometry of biomass partitioning and allocation can explain interspecific competition (Yu & Gao, 2011).
According to the nutrition tolerance/resistance and morphology-related allocation strategy (Yu & Gao, 2011), S. krylovii stimulated aboveground biomass to grow higher and increase light interception, while A. frigida stimulated belowground biomass to acquire soil nutrients or stimulated stolon biomass to some extent to colonize new patches in our study. Therefore, we considered that S. krylovii increased competitive ability through stimulated growth and ability to affect neighbors, whereas A. frigida increased competitive ability by stimulating the ability to avoid being affected (Kolodziejek, 2019;Wang et al., 2010).

| The maternal environment affected the species' interspecific relationship
Our results are in line with studies which show that the maternal environment may alter offspring performance (Guillaume et al., 2016).
However, they were not in line with a previous study showing that perennial plants showed hardly any transgenerational responses (Yin et al., 2019). This is because most of the perennial plants included in that meta-analysis were trees, whereas those in our study were grass or subshrubs which are more vulnerable to environmental changes (Walter et al., 2016). Moreover, perennial plants may benefit from the transgenerational effect as adaptation is unlikely to keep pace with rapid environmental changes (Herman & Sultan, 2011).
The transgenerational effect of maternal environment on offspring may be attributed to maternal plant phenotypic plasticity to their environment. Species exhibit different phenotypic plasticities to their immediate surroundings via their own physiological optimum (Gilman et al., 2010;Hoffmann & Sgrò, 2011). Thus, the transgenerational effects of the two species in our study were species-specific, even when maternal plants experienced the same environment. As mentioned above, as a superior competitor, S. krylovii invested more nutrients in shoots to acquire light to maintain competitive advantage under favorable immediate conditions. As water addition in this sterile environment only improved limited nutrient availability, the maternal plant of S. krylovii might not acquire enough nutrients from the soil to maintain survival, growth, and fecundity (investment in offspring) concurrently, and thus had little nutrient invested in offspring . Thus, offspring biomass may decrease with maternal water addition. However, A. frigida, as a weaker competitor, invested equally in biomass for the shoots and the roots to acquire enough soil nutrients under favorable immediate conditions. As the maternal plant of A. frigida had accumulated more nutrients to invest in offspring, offspring biomass increased with maternal nitrogen and water addition. Our results were also consistent with previous evidence indicating that lower resource acquisition reduced maternal resource allocation to offspring (Stotz et al., 2018) and that larger plants do not increase offspring performance proportionately (Aarssen, 2015).
Besides the offspring trait, maternal environment affected the offspring competitive ability according to our results and to that of previous studies (Stotz et al., 2018). The effects of the maternal environment and immediate environment may be counteractive (Lau et al., 2008). For example, according to our results, unlike the immediate environment's effect, the transgenerational effect of maternal environment decreased S. krylovii's competitive ability The specific transgenerational effect on the competitive ability of the two species may be attributed to different population development strategies. In order to maximize population fitness, annual plants improve population fitness mainly by increasing investment to offspring (Neytcheva & Aarssen, 2008), whereas perennial plants have more tradeoffs (Yin et al., 2019). To avert the deleterious consequences of environmental change, immediate action must be taken for the survival and growth of perennial plants (Auge et al., 2017;Suzuki & Teranishi, 2005;Tracey & Aarssen, 2011, 2014, and strategic conservation planning for the coming years and offspring is necessary (Dawson et al., 2011).
That is, plant populations exhibit phenotypic plasticity to cope with environmental changes either in the generation or in filial generations (Holeski et al., 2012;Van Dam & Baldwin, 2001).
There are no general apparent advantageous strategies for population development. For example, if the environment fluctuates at a high frequency, within-generational plasticity will be advantageous for superior competitors or species with rapid plasticity growth (Auge et al., 2017;Kuijper & Hoyle, 2015), while weaker competitors relieve competitive suppression through avoidance , which allocate more resources to offspring instead of growth (Aarssen, 2015). Thus, although the direction and magnitude may be different between the maternal environment effect and immediate environment effect for the two species in our study, we cannot determine whether the transgenerational effect is adaptive based on few offspring traits, especially for perennial plants. As a plant population consists of individuals of different generations, assessing the adaptive value of transgenerational effect on one generation's fitness is insufficient (Auge et al., 2017;Uller, 2008).

| The interaction of N addition and increased precipitation
Water and nitrogen are the most frequent limiting resources in temperature steppes (Xu et al., 2018). Changes in nitrogen and water availability are known to influence vegetation dynamics and ecosystem processes . Water is essential for nutrient diffusion and replenishment in soil (Zhang et al., 2015). High water availability may contribute to the replenishment of added N to the soil solution and consequently increase available nitrogen for plant growth. Thus, in arid and sterile environments, water and nitrogen addition usually play a significant interaction. As a superior competitor, S. krylovii used the increase in nitrogen availability to stimulate its biomass in the mixture in an immediate nitrogen and water coinstantaneous added environment. This may explain the significant interaction between nitrogen addition and water addition in the competitive responses of S. krylovii in our study.
However, S. krylovii could not transfer competitive advantage to its filial generation because it took advantage of favorable immediate environments to promote competitive ability. Although the F I G U R E 3 Maternal-treatment-induced changes in biomass, aboveground biomass, and belowground biomass of the two species grow in monoculture and in mixture. Values are means ± SE, n = 6 for each treatment. S.k: Stipa krylovii, A.f: Artemisia frigida. M-CK: control treatment; M-N: nitrogen addition in Field Experiment (maternal N addition environment); M-Pr: water addition Field Experiment (maternalincreased precipitation environment); M-NPr: nitrogen addition and water addition in combine in field experiment (maternal N addition plus increased precipitation environment). Different uppercase letters indicate significant differences among treatments for S. krylovii or A. frigida grown in monoculture; different lowercase letters indicate significant differences among treatments for S. krylovii or A. frigida grown in mixture F I G U R E 4 Maternal-treatment-induced changes of competitive effect. Values are means ± SE, n = 6 for each treatment. See Figure 3 for abbreviations. Different uppercase letters indicate significant differences among treatments for S. krylovii; different lowercase letters indicate significant differences among treatments for A. frigida maternal nitrogen and water coinstantaneous added environment substantially stimulated the biomass of A. frigida in the monoculture, as a weaker competitor, A. frigida in the mixture stimulated little biomass and consequently exhibited a lower CR. The interaction of water addition and nitrogen addition to arid and sterile environments showed that S. krylovii stimulated competitive ability in the generation and A. frigida stimulated biomass accumulation in the filial generation to some extent. Therefore, it is more important to take the transgenerational effect on the interspecific competition of offspring into account under future increased precipitation and nitrogen addition.

| CON CLUS IONS
Our results suggest that species

CO N FLI C T O F I NTE R E S T
The authors declare that they have no conflict of interest.

DATA AVA I L A B I L I T Y S TAT E M E N T
The data that support the findings of this study are openly available in the Dryad Digital Repository at https://doi.org/10.5061/dryad. Ming Yue https://orcid.org/0000-0002-9092-0251