Small-scale plant species distribution in snowbeds and its sensitivity to climate change

Alpine snowbeds are characterized by a long-lasting snow cover and low soil temperature during the growing season. Both these key abiotic factors controlling plant life in snowbeds are sensitive to anthropogenic climate change and will alter the environmental conditions in snowbeds to a considerable extent until the end of this century. In order to name winners and losers of climate change among the plant species inhabiting snowbeds, we analyzed the small-scale species distribution along the snowmelt and soil temperature gradients within alpine snowbeds in the Swiss Alps. The results show that the date of snowmelt and soil temperature were relevant abiotic factors for small-scale vegetation patterns within alpine snowbed communities. Species richness in snowbeds was reduced to about 50% along the environmental gradients towards later snowmelt date or lower daily maximum temperature. Furthermore, the occurrence pattern of the species along the snowmelt gradient allowed the establishment of five species categories with different predictions of their distribution in a warmer world. The dominants increased their relative cover with later snowmelt date and will, therefore, lose abundance due to climate change, but resist complete disappearance from the snowbeds. The indifferents and the transients increased in species number and relative cover with higher temperature and will profit from climate warming. The snowbed specialists will be the most suffering species due to the loss of their habitats as a consequence of earlier snowmelt dates in the future and will be replaced by the avoiders of late-snowmelt sites. These forthcoming profiteers will take advantage from an increasing number of suitable habitats due to an earlier start of the growing season and increased temperature. Therefore, the characteristic snowbed vegetation will change to a vegetation unit dominated by alpine grassland species. The study highlights the vulnerability of the established snowbed vegetation to climate change and requires further studies particularly about the role of biotic interactions in the predicted invasion and replacement process.     

Advanced snowmelt causes shift towards positive neighbour interactions in a subarctic tundra community

Positive and negative species interactions are important factors in structuring vegetation communities. Studies in many ecosystems have focussed on competition; however, facilitation has often been found to outweigh competition under harsh environmental conditions. The balance between positive and negative species interactions is known to shift along spatial, temporal and environmental gradients and thus is likely to be affected by climate change. Winter temperature and precipitation patterns in Interior Alaska are rapidly changing and could lead to warmer winters with a shallow, early melting snow cover in the near future. We conducted snow manipulation and neighbour removal experiments to test whether the relative importance of positive and negative species interactions differs between three winter climate scenarios in a subarctic tundra community. In plots with ambient, manually advanced or delayed snowmelt, we assessed the relative importance of neighbours for survival, phenology, growth and reproduction of two dwarf shrub species. Under ambient conditions and after delayed snowmelt, positive and negative neighbour effects were generally balanced, but when snowmelt was advanced we found overall facilitative neighbour effects on survival, phenology, growth and reproduction of Empetrum nigrum, the earlier developing of the two target species. As earlier snowmelt was correlated with colder spring temperatures and a higher number of frosts, we conclude that plants experienced harsher environmental conditions after early snowmelt and that neighbours could have played an important role in ameliorating the physical environment at the beginning of the growing season.     

Experimental comparison of competition and facilitation in alpine communities varying in productivity

Question. Competitive and facilitative interactions among plant species in different abiotic environments potentially link productivity, vegetation structure, species composition and functional diversity. We investigated these interactions among four alpine communities along an environmental productivity gradient in a generally harsh climate. We hypothesised that the importance of competition would be higher in more productive sites. Location. Mt. M. Khatipara (43°27′N, 41°41′E, altitude 2750 m), NW Caucasus, Russia. Communities ranged from low‐productivity alpine lichen heath (ALH) and snowbed communities (SBC), to intermediate productivity Festuca grassland (FVG), and high‐productivity Geranium‐Hedysarum meadow (GHM). Methods. We quantified the relative influence of competition and facilitation on community structure by expressing biomass of target species within each natural community proportionally to biomass of the species in a “null community” with experimental release from interspecific competition by removing all other species (for 6 years). An overall index of change in community composition due to interspecific interactions was calculated as the sum of absolute or proportional differences of the component species. Results. Species responses to neighbour removal ranged from positive to neutral. There was no evidence of facilitation among the selected dominant species. As expected, competition was generally most important in the most productive alpine community (GHM). The intermediate position for low‐productivity communities of stressful environments (ALH, SBC) and the last position of intermediately productive FVG were unexpected. Conclusions. Our results appear to support the Fretwell‐Oksanen hypothesis in that competition in communities of intermediate productivity was less intense than in low‐ or high‐productive communities. However, the zero net effect of competition and facilitation in FVG might be the result of abiotic stress due to strong sun exposure and high soil temperatures after neighbour removal. Thus, non‐linear relationships between soil fertility, productivity and different abiotic stresses may also determine the balance between competition and facilitation.     

No increase in alpine snowbed productivity in response to experimental lengthening of the growing season

Climate change effects on snow cover and thermic regime in alpine tundra might lead to a longer growing season, but could also increase risks to plants from spring frost events. Alpine snowbeds, i.e. alpine tundra from late snowmelt sites, might be particularly susceptible to such climatic changes. Snowbed communities were grown in large monoliths for two consecutive years, under different manipulated snow cover treatments, to test for effects of early (E) and late (L) snowmelt on dominant species growth, plant functional traits, leaf area index (LAI) and aboveground productivity. Spring snow cover was reduced to assess the sensitivity of snowbed alpine species to severe early frost events, and dominant species freezing temperatures were measured. Aboveground biomass, productivity, LAI and dominant species growth did not increase significantly in E compared to L treatments, indicating inability to respond to an extended growing season. Edapho‐climatic conditions could not account for these results, suggesting that developmental constraints are important in controlling snowbed plant growth. Impaired productivity was only detected when harsher and more frequent frost events were experimentally induced by early snowmelt. These conditions exposed plants to spring frosts, reaching temperatures consistent with the estimated freezing points of the dominant species (～?10 °C). We conclude that weak plasticity in phenological response and potential detrimental effects of early frosts explain why alpine tundra from snowbeds is not expected to benefit from increased growing season length.     

Release from competition and protection determine the outcome of plant interactions along a grazing gradient

Plant interactions are suggested to shift from competition to facilitation and collapse with increasing grazing pressure. The existence of this full range of plant interactions and the role of underlying mechanisms (i.e. release from competition and protecting effect) in response to herbivory remains poorly documented and mainly described in terrestrial systems. We use a large grazing disturbance gradient (five levels of grazing) to test its effect on the outcome of plant interactions and underlying mechanisms in freshwater ecosystems. In a mesocosm experiment, we manipulated the presence of neighbouring plants to test their negative (competition) or protective (facilitation) effects on target plants along the grazing pressure gradient. We predicted that plant interactions 1) shift from competition to indirect facilitation with increased grazing pressure, 2) indirect facilitation collapses at high levels of grazing, 3) release from competition mainly drives the outcome in lowly grazed conditions and, 4) decreased protection occurs in highly grazed conditions responsible for the collapse of facilitation. This study shows the occurrence of the full range of outcomes in plant interactions under a wide spectrum of grazing pressure and indicates how the complex combination of underlying mechanisms shapes variations in plant interactions. We show that both, the release from competition and the increased protection by neighbouring plants drove the shift from competition to indirect facilitation. Declined protection by neighbouring plants resulted in a collapse of indirect facilitation for survival under intense herbivory. Our study provides the first experimental evidence of indirect facilitation structuring freshwater ecosystems thereby validating important ecological concepts mainly developed for terrestrial ecosystems.     

Friends or Foes? Interplay of facilitation and competition depends on the interaction between abiotic stress and ontogenetic stage

The importance of ontogeny and the degree of abiotic stress in determining the interplay between facilitation and competition is well known. However, their joint effect on the outcome of plant interactions remains poorly understood, especially when a continuous gradient of abiotic stress is considered. Our objective was to evaluate the frequency of association of individuals of Clusia criuva with typical coastal dune species across a gradient of water stress and how this association affects the growth of juveniles and sub-adults. The study was performed in a coastal dune region in South Brazil, where the sandy soil promotes severe water stress. One-year growth of 293 individuals and their distance to the closest humid slacks were measured. This distance is a good surrogate for water stress, since slacks represent proximity to groundwater. The proportion of associated individuals increased with abiotic stress in both ontogenetic stages, but was always greater for juveniles. This suggests that association is progressively more important to guarantee survival as abiotic stress increases. Nonetheless, the benefit of neighbors to growth decreased with abiotic stress, and associated plants grew less than isolated ones in harsher environments. This was mainly true for juveniles, since the height growth of sub-adults was not affected by association or abiotic stress. In our study, facilitation became more intense with environmental severity, increasing survival, although competition also became more influent, reducing growth particularly for younger plants. This demonstrates that ontogenetic stage and abiotic stress must be considered simultaneously in order to better understand interactions among plants.     

Facilitation and interference in subalpine meadows of the central Caucasus

Abstract. We studied the effects of neighbours on the biomass of seven randomly chosen species in species‐rich sub‐alpine meadows in the central Caucasus Mountains by comparing the performance of plants with neighbours removed experimentally to that of paired plants with their neighbours left intact. In most cases the removal of neighbours led to significant increases in vegetative and total above‐ground biomass implying the species were limited by competition. However, the neighbour removal led also to an increased leaf wilting for target plants, as well as to strong decline in reproductive effort for some species. We hypothesise that competition may be the prevailing type of interaction in species‐rich sub‐alpine meadow communities, but competitive effects on vegetative production may be balanced, if not outweighed, by facilitation, at least for some species. Such a balance may enhance species coexistence in communities.     

Simulated herbivory counteracts the effects of grass competition on the invasive fireweed (Senecio madagascariensis)

Competition, herbivory and their interaction play a significant role in determining the competitive ability and survival of individual plant species. Understanding these processes and interactions can improve the efficacy of biocontrol programs against invasive weeds. Senecio madagascariensis (fireweed) is an invasive weed of South African origin that reduces pastoral productivity and poisons livestock in several countries, notably Australia. Although competitive pastures can suppress the weed’s growth in Australia, its competitive nature is poorly understood in relation to its invasion success. This greenhouse study assessed the growth and reproductive yield of fireweed growing in competition with six native and introduced grasses present in both South Africa and Australia. Since fireweed is a target for biocontrol in Australia, we examined whether its response to grass competition changed with herbivory (simulated by 40% leaf removal). The effect of grass competition and herbivory on the weed’s biomass and floral productivity was examined during a 12‐week pot trial in South Africa. Floral numbers were unaffected by both grass competition and herbivory. Biomass was used to calculate Relative Interaction Indices (RII) to quantify the weed’s competitive or facilitative response. This index compares a specific measurable trait, such as biomass, of fireweed growing alone, to fireweed growing with grass to determine the level of competitive suppression or facilitation resulting from the interaction. Despite the lack of species‐specific effects of grass competition, the presence of grass suppressed fireweed’s foliar, root and whole plant biomass the most when herbivory was absent. With herbivory, fireweed did not suffer from any measurable competitive suppression. This lack of competitive suppression may be due to an induced allelopathic response, given the levels of pyrrolizidine alkaloids common in many Senecio species. Since this result may weaken the case for biocontrol, the weed’s competitive responses should be verified in relation to actual insect herbivory.     

Context‐dependent biotic interactions control plant abundance across altitudinal environmental gradients

Many biotic interactions influence community structure, yet most distribution models for plants have focused on plant competition or used only abiotic variables to predict plant abundance. Furthermore, biotic interactions are commonly context‐dependent across abiotic gradients. For example, plant–plant interactions can grade from competition to facilitation over temperature gradients. We used a hierarchical Bayesian framework to predict the abundances of 12 plant species across a mountain landscape and test hypotheses on the context‐dependency of biotic interactions over abiotic gradients. We combined field‐based estimates of six biotic interactions (foliar herbivory and pathogen damage, fungal root colonization, fossorial mammal disturbance, plant cover and plant diversity) with abiotic data on climate and soil depth, nutrients and moisture. All biotic interactions were significantly context‐dependent along temperature gradients. Results supported the stress gradient hypothesis: as abiotic stress increased, the strength or direction of the relationship between biotic variables and plant abundance generally switched from negative (suggesting suppressed plant abundance) to positive (suggesting facilitation/mutualism). For half of the species, plant cover was the best predictor of abundance, suggesting that the prior focus on plant–plant interactions is well‐justified. Explicitly incorporating the context‐dependency of biotic interactions generated novel hypotheses about drivers of plant abundance across abiotic gradients and may improve the accuracy of niche models.     

The interplay between above‐ and below‐ground plant–plant interactions along an environmental gradient: insights from two‐layer zone‐of‐influence models

The changes in plant–plant interactions along environmental gradients have been a focus of recent ecological research. It has been suggested that both above‐ and below‐ground competition and their interplay vary along gradients, but few studies have investigated this idea, and in most cases, the role of facilitation has not been considered, despite its importance in high stress environments. Here we used two‐layer ‘zone‐of‐influence’ models to simulate the effects of facilitation, size‐asymmetry of competition, abiotic stress, resource availability and the balance of root–shoot growth on shoot and root interactions and their interplay along an environmental gradient. In the absence of facilitation, shoot and total competition became weaker, while root competition and the interplay between shoot and root competition were unchanged under increasing stress when root competition was completely symmetric. In contrast, shoot, root, total interactions and the interplay between shoot and root interactions were all negative, and they increased with increasing stress when root competition was size‐symmetric. When facilitation was included in the models, net effects of shoot, root, total interactions and the interplay of root–shoot interactions were very different from those without facilitation, and many were positive under highly stressful conditions. The type of stress (non‐resource or resource) did not significantly influence the simulation results. Our study provides an alternative interpretation of the interplay between above‐ and below‐ground plant–plant interactions across an environmental gradient.     

Disentangling Competition, Herbivory, and Seasonal Effects on Young Plants in Newly Restored Communities

Optimizing techniques of impact and consequence assessment are critical when faced with the challenges of reclamation within a damaged or altered ecosystem. Much debate has arisen over an appropriate index to evaluate herbivore and competition effects on restored communities. We assessed concurrent environmental pressures by means of repeated measurements using three common indices of plant performance (biomass, shoot extension, and survival) in conjunction with monitoring for number and timing of plants eaten. Our design incorporated 24 species, representing a range of taxonomic groups and growth forms, planted at low and high densities, inside and outside large‐scale mammal exclosures. We demonstrate that biomass and height measurements are correlated (at both the individual and the combined species levels), whereas the survival index often showed independent information. Using the most conservative measure (survival), we delineate between plant deaths attributed to seasonal effects, competition (some facilitation was apparent), and herbivory (both compensation and loss of fitness were demonstrated). Plant spacing effects depended on the index (response variable) and whether we measured individual or combined species. The survival index rarely showed competition effects. Due to counter facilitation effects, competition was not demonstrated for any index at the combined species level. The comparison of the relative order and magnitude of plants being eaten against impact identified vulnerable and compensating species. Once identified, compensating species may be used sacrificially to buffer damage in new reclamation systems, whereas deterrents may be used around known vulnerable species.     

Additive and nonadditive effects of herbivory and competition on tree seedling mortality, growth, and allocation

The interaction between simulated cotyledon herbivory and interspecific competition was studied in a greenhouse experiment using two species of trees, Acer rubrum and Quercus palustris, which commonly invade abandoned agricultural fields. Herbivory treatments were applied as a gradient of cotyledon removal for A. rubrum with 0, 25, 50, 75, and 100% of cotyledon tissue removed. Cotyledons from Q. palustris were clipped and removed (control, early, and late removal) to create a gradient of seed reserve availability. The competition treatment consisted of plugs of old-field vegetation that filled the pots with perennial cover. Mortality of seedlings was higher with competition. There was a significant interaction between herbivory and competition with the highest mortality occurring with competition at the highest intensity of herbivory in both species. Herbivory reduced biomass for Q. palustris only, while competition reduced biomass in both species. Neither species showed an interaction between herbivory and competition for growth. There was a significant interaction between herbivory and competition on allocation patterns for both species, with greater allocation to roots with competition at the highest intensity of herbivory. This study demonstrates the potential for cotyledon herbivory and competition to interact, altering the invasion of tree seedlings into abandoned agricultural land.     

Changes in plant interactions along a gradient of environmental stress

A combination of competition and facilitation effects operating simultaneously among plant species appears to be the rule in nature, where these effects change along productivity gradients often in a non-proportional manner. We investigated changes in competition and facilitation between a leguminous shrub, Retama sphaerocarpa , and its associate understorey species along an environmental gradient in semi-arid southeast Spain. Our results show a change in the net balance of the interaction between the shrub and several of its associated species, from clearly positive in the water-stressed, infertile environment to neutral or even negative in the more fertile habitat. There was a weakening of facilitation along the fertility gradient as a consequence of improved abiotic conditions. Competition was the most intense for below-ground resources in the less fertile environment while total competition tended to increase towards the more productive end of the gradient. Changes in the balance of the interaction between and among different plant species along the gradient of stress were caused by a decline in facilitation rather than by a change in competition. As both competition intensity and facilitation change along gradients of resource availability, plant interactions are best viewed as dynamic relationships, the outcome of which depends on abiotic conditions.     

Interactive effects of herbivory and competition intensity determine invasive plant performance

Herbivory can reduce plant fitness, and its effects can be increased by competition. Though numerous studies have examined the joint effects of herbivores and competitors on plant performance, these interactive effects are seldom considered in the context of plant invasions. Here, we examined variation in plant performance within a competitive environment in response to both specialist and generalist herbivores using Chinese tallow as a model species. We combined tallow plants from native and invasive populations to form all possible pairwise combinations, and designated invasive populations as stronger neighbours and native populations as weaker neighbours. We found that when no herbivory was imposed, invasive populations always had higher total biomass than natives, regardless of their neighbours, which is consistent with our assumption of increased competitive ability. Defoliation by either generalist or specialist herbivores suppressed plant growth but the effects of specialists were generally stronger for invasive populations. Invasive populations had their lowest biomass when fed upon by specialists while simultaneously competing with stronger neighbours. The root/shoot ratios of invasive populations were lower than those of native populations under almost all conditions, and invasive plants were taller than native plants overall, especially when herbivores were present, suggesting that invasive populations may adopt an "aboveground first" strategy to cope with herbivory and competition. These results suggest that release from herbivores, especially specialists, improves an invader's performance and helps to increase its competitive ability. Therefore, increasing interspecific competition intensity by planting a stronger neighbour while simultaneously releasing a specialist herbivore may be an especially effective method of managing invasive plants.     

Sources of variation in plant responses to belowground insect herbivory: a meta-analysis

Growing interest in belowground herbivory and the remarkable diversity of the accumulated information on this topic inspired us to quantitatively explore the variation in the outcomes of individual studies. We conducted a meta-analysis of 85 experimental studies reporting the effects of root-feeding insect herbivores (36 species) on plants (75 species). On average, belowground herbivory led to a 36.3% loss of root biomass, which was accompanied by a reduction in aboveground growth (-16.3%), photosynthesis (-11.7%) and reproduction (-15.5%). The effects of root herbivory on aboveground plant characteristics were significant in agricultural and biological control studies, but not in studies of natural systems. Experiments conducted in controlled environments yielded larger effects on plants than field experiments, and infestation experiments resulted in more severe effects than removal studies employing natural levels of herbivory. Simulated root herbivory led to greater aboveground growth reductions than similar root loss imposed by insect feeding. External root chewers caused stronger detrimental effects than sap feeders or root borers; specialist herbivores imposed milder adverse effects on plants than generalists. Woody plants suffered from root herbivory more than herbaceous plants, although root loss was similar in these two groups. Evergreen woody plants responded to root herbivory more strongly than deciduous woody plants, and grasses suffered from root herbivory more than herbs. Environmental factors such as drought, poor nutrient supply, among-plant competition, and aboveground herbivory increased the adverse effects of root damage on plants in an additive manner. In general, plant tolerance to root herbivores is lower than tolerance to defoliating aboveground herbivores.     

Site-specific factors influence the richness and phenology of snowbed plants in the Pyrenees

Although the timing of snowmelt and growth temperatures appear to be the main factors that influence the species richness and phenology of snowbed plants, site-specific characteristics may also play a role in modifying the effects of the timing of snowmelt and temperature. In this study, the effects of site-specific factors (microtopography and snow origin) on species richness and plant phenology were evaluated in 72 plots in two snowbeds in the Andorran Pyrenees. Snowmelt patterns influenced the spatial distribution of species richness and abundance. Site-specific factors had significant effects on the responses of species (shortening or lengthening the duration of the phenophase) and on the extent to which the timing of snowmelt influenced leaf expansion and flowering. Notably, the highest rates of leaf expansion occurred on late snowmelt isoclines, where, nevertheless, the time taken to reach peak flowering was significantly longer than on the early snowmelt isoclines. The results of this study highlight the fact that, in addition to the effects of interannual variability in climate, site-specific factors have a significant effect on the phenology and reproductive success of the commonest plants in the snowbed communities of the Pyrenees.     

No interaction between competition and herbivory in limiting introduced Cirsium vulgare rosette growth and reproduction

Both competition and herbivory have been shown to reduce plant survival, growth, and reproduction. Much less is known about whether competition and herbivory interact in determining plant performance, especially for introduced, weedy plant species in the invaded habitat. We simultaneously evaluated both the main and interactive effects of plant neighbors and insect herbivory on rosette growth and seed reproduction in the year of flowering for Cirsium vulgare (bull thistle, spear thistle), an introduced Eurasian species, in tallgrass prairie in 2 years. Effects of insect herbivory were strong and consistent in both years, causing reduced plant growth and seed production, whereas the effects of competition with established vegetation were weak. The amount of herbivore damage inflicted on rosettes did not depend on the presence of neighbor plants. We also found no interaction between competition and herbivory on key parameters of plant growth and fitness. The results of this study contradict the hypothesis that competitive context interacts with insect herbivory in limiting the invasiveness of this introduced thistle. Further, the results provide additional, experimental evidence that high levels of herbivory on established rosettes by native insects exert significant biotic resistance to the invasiveness of C. vulgare in western tallgrass prairie.     

Herbivory and plant resource competition: a review of two interacting interactions

This review discusses the prevalence and potential for interactive effects between herbivory and competition on plant growth and biomass, and it is apparent that such effects typically arise when there is a mismatch between the spatial scale of herbivore behaviour (food or patch choice) and the spatial heterogeneity of the plant community. Historically, such interactive effects have been examined using two approaches. Studies using the first approach have excluded plant neighbors and herbivores in a factorial experiment, and scored effects on plant biomass. Studies using the second approach have observed herbivore abundance or herbivory on plants with or without plant neighbors, and have identified a large number of mechanisms underlying such interactive effects. The two types of studies have produced somewhat conflicting results, where interactive effects have been commonly observed in studies using the second approach and only rarely in studies using the first approach. This is most likely a consequence of a biased choice of study systems, where studies using the first approach have primarily studied mammalian herbivory while studies using the second approach have been more focussed on insect herbivory. Moreover, studies using the first approach have typically been very small-scale manipulations and this probably precludes most possible interactive effects in systems with mammalian herbivory. This points to the fact that studies examining interactive effects of herbivory and plant competition should more carefully consider the behaviour and life history of herbivores included in the study prior to the design of removal experiments.     

The importance of facilitation in the zonation of shrubs along a coastal salinity gradient

Question: What are the interactive roles of abiotic stress and plant interactions in mediating the zonation of the shrub Tamarix chinensis along a salinity gradient? Location: Yellow River estuary (37°46′N, 119°09′E), northeast China. Methods: We surveyed the zonation of T. chinensis along a salinity gradient and quantified its salt tolerance using a pot experiment. In two field experiments, we transplanted T. chinensis seedlings into salt marsh, transitional zone and upland habitats, manipulated neighbours and quantified survivorship and biomass to examine neighbour effects. We also quantified vegetation effects on abiotic conditions in each zone. Results: Tamarix chinensis dominated the transitional zone, but was absent in upland and salt marsh habitats. In the pot experiment, T. chinensis grew well in freshwater treatments, but was inhibited by increasing salinity. Field experiments revealed that competition from neighbours limited T. chinensis growth in the uplands, while T. chinensis transplants were limited, with or without neighbours, in the salt marsh by high soil salinity. In the transitional zone, however, T. chinensis transplants performed better with than without neighbours. Vegetation removal significantly elevated soil salinity in the transitional zone, but not in other zones. Conclusions: Competition, facilitation and abiotic stress are all important in mediating the zonation of T. chinensis. Within its physiological stress tolerance range, or fundamental niche, it is limited by plant competition in low salinity habitats, and facilitated by neighbours in high salt stress habitats, but cannot survive in salt marshes having salinities above its salt stress tolerance limit. Our results have implications for understanding the relationships between facilitation and stress gradients.     

Extreme drought stress shifts net facilitation to neutral interactions between shrubs and sub‐canopy plants in an arid desert

The stress gradient hypothesis (SGH) predicts that the importance or intensity of competition and facilitation will change inversely along abiotic stress gradients. It was originally postulated that increasing environmental stress can induce a monotonic increase in facilitation. However, more recent models predicted that the relationship between severity and interaction exhibits a hump‐shaped pattern, in which positive interactions prevail under moderate stress but decline at the extreme ends of stress gradients. In the present study, we conducted a field experiment along a temporal rainfall gradient for five consecutive years, in order to investigate interactions in a shrub‐herbaceous plant community at the southern edge of the Badain Jaran Desert, and, more specifically, investigated the effects of Calligonum mongolicum, a dominant shrub species, on both abiotic environmental variables and the performance of sub‐canopy plant species. We found that shrubs can improve sub‐canopy water regimes, soil properties, plant biomass, density, cover, and richness and, more importantly, that the positive effect of shrubs on sub‐canopy soil moisture during the summer diminishes as rainfall decreases, a pattern that partly explains the collapse of the positive interaction between shrubs and their understory plants. These results provide empirical evidence that the positive effect of shrubs on understory plant communities in extreme arid environments may decline and become neutral with increasing drought stress.