Predicting Species Response to Changes in Resource Availability

Predicting Species Response to Changes in Resource Availability

Report of the Fertilization Synthesis Group from the 2003 ASM

Christopher Clark, CDR

Scott Collins, SEV and KNZ

Laura Gough, ARC

Katherine Gross, KBS

Daniel Milchunas, SGS

Steven Pennings, GCE

Katharine Suding, NWT

The Grand Challenge. Human modification of climate and land-use changes have increased global terrestrial net primary production 6% over the past two decades (Nemani et al. 2003). Human activity now fixes more atmospheric N₂ into biologically-available forms of nitrogen than all natural terrestrial processes combined (Vitousek et al. 1997). Deposition of N from air pollution has increased 5-20 times in many areas relative to pristine conditions. Increased nitrogen availability will increase primary productivity because of the nitrogen limitation of many terrestrial and marine ecosystems (Vitousek and Howarth 1991). In addition, climate changes such as increased temperature and precipitation increase the rate of microbial transformations of nitrogen into plant-available forms (Vitousek et al. 1997). Humans have also substantially modified the availability of phosphorus (particularly in aquatic systems), water, and other critical resources. Ecological feedbacks to increased resource availability have the potential to affect the dynamics of vector-borne diseases (Wolfe and Patz 2002), increase allergenic pollen production (Townsend et al. 2003) and enhance the success of problematic invasive species (Davis et al. 2000).

The intermediary that likely drives these feedbacks in terrestrial ecosystems is plant species composition and diversity. Thus, understanding how species respond to increased resource availability and how this response influences species composition and abundance may be the key to predicting the severity and urgency of these feedbacks to human health, the economy, and system stability. Unfortunately, while many mechanisms have been proposed to determine patterns of species diversity along productivity gradients (Mittelbach et al 2001), which of these mechanisms operate at different spatial scales and how they influence community structure remain puzzling. Thus our ability to forecast how communities will respond to changes in external and internal factors influencing local and regional diversity and function remain weak. As landscapes continue to be dramatically altered, we will need coordinated research to develop a broad-based understanding of mechanisms that create, maintain, and decrease taxonomic and functional diversity at multiple scales.

To make progress in this critical area of research, we must develop a better understanding of the relationship between aboveground primary productivity, resource fluctuations and species diversity. Over the past three decades, this relationship has received considerable experimental and observational attention (including a past LTER synthesis effort: Waide et al. 1999, Gross et al. 2000, Gough et al. 2000, Mittelbach et al. 2001), and has been used to theoretically and empirically link community and ecosystem attributes. However, the mechanisms that drive the relationship between productivity and diversity remain uncertain (Abrams 1995, Oksanen 1996, Stevens and Carson 2002). Thus, our current state of knowledge is hampering our ability to forecast the causes and consequences of change due to human-modifications of resource availability.

While plant diversity change across productivity gradients is quantified almost entirely by changes in the distribution and abundance of taxonomic richness, system functioning and ecological feedbacks may be more closely tied to changes in functional richness and dominance. Hence, one approach towards understanding diversity changes in response to changes in productivity is to examine the types of species that are lost and those that increase in dominance following experimental manipulations of resources.

LTER Network Involvement. The LTER Network has a tremendous resource of on-going, long-term experiments that are perfectly suited to address questions such as the ones outlined above. These experiments include direct addition of nitrogen, water, and/or phosphorus, changes in the disturbance regime through grazing, fire, and clearcutting, and environmental alteration with warming chambers, snowfences, and carbon dioxide. In virtually all these experiments, plant species response (and at times responses at other trophic levels) is measured over time, both in terms of taxonomic diversity and abundance patterns.

A group of LTER investigators from eight sites have begun a synthesis effort to explore further the wealth of data available from nitrogen fertilization experiments conducted in herbaceous, short-stature plant communities. We came together to follow up on a 1996 NCEAS workshop focused on the relationship between productivity and diversity at LTER sites (organized by R. Waide and M. Willig). In 2002, we began to construct a database of fertilization experiments across eight ecosystems and twenty-two community types involving the experiments we examined in our previous synthesis effort (see citations above). The responses (relative abundance in control and nitrogen addition plots) of 830 species records are currently included. We have described each species according to a suite of traits: growth form, life history, relative height, clonality, and origin (native, non-native). In addition, each community type is described according to parameters that may influence functional response: system productivity, species pool, and climatic variables. We plan to expand this database and then use a variety of analytical approaches (structural equation modeling, null modeling, meta-analysis) to address whether there are generalities about which functional groups respond to fertilization and how environmental factors interact with these responses.

At the 2003 ASM we conducted a workshop to present some of our first analyses of this large dataset to the LTER community. Initial examination of species responses across all sites suggests that more species are lost (27%) than gained (16%) in nitrogen-addition plots when compared with control plots. However, the number of common species that increased in relative abundance with added nitrogen was similar to the number that decreased (18% and 21%, respectively), suggesting there might be compensation in relative abundance by the common species. We have begun examining effects of treatments on environmental variables as well as correlations between community parameters and species response. For example, surprisingly, more productive sites showed a greater proportional decrease in light availability with added nitrogen than lower productivity sites. Native species appeared to increase more with nitrogen fertilization at warmer sites, but performed less well at cooler sites, while the opposite pattern was true of non-native species. This finding may help us better predict invisibility of communities across sites. We also presented results related to particular species traits discussing which traits were correlated with extinction or increased dominance with added resources, and a first look at whether or not individual plant species behaved similarly in response to the same manipulation within and across sites. These examples represent just a small set of the questions we are currently investigating.

Table 1. Important and cross-cutting questions that could be addressed by this project. We expect these questions to evolve with time and interactions among participants.

Species Loss. What types of species are lost when diversity declines due to enhanced resource availability? Does diversity change more in sites where an abundant species is lost? What shifts occur when only rare species are lost? In what type of system does each type of loss occur?

Dominance Shift. What are the characteristics of species that become dominant following fertilization? Does increased resource availability encourage invasion by exotic species? Is morphological or physiological plasticity related to fertilization response? Are these characteristics consistent with those predicted by plant competition theory? What types of environments are more at risk of a species “taking over” following increased resource availability?

Temporal Change. How do short-term responses differ from longer-term responses? When does immigration play a role in species change? Does species response to fertilization or change along gradients predict how systems will respond to general environmental change?

Spatial Variation. How does change at one spatial scale relate to other spatial scales? Does within-site or within-region heterogeneity influence these relationships? How does variation in available species (i.e., species pool), regardless of traits, affect diversity? Are other factors that often co-vary with productivity (such as heterogeneity or land-use history) better predictors of diversity and species change?

In the near future we envision initiating several cross-site experiments using standardized methodologies to test relationships detected in our current dataset. For example, we might ask if experimental reduction in resources reverses fertilization-induced changes in diversity and species composition. Scant evidence is available on species responses to reductions in productivity. At Konza Prairie, however, Knapp et al. (2002) found that ANPP decreased and species diversity increased under altered precipitation regimes, and Baer et al. (2003) found that diversity of native species was highest and abundance of non-native species was lowest on carbon-amended soils (with lower nitrogen) during the early stages of prairie restoration. Second, can we generalize beyond increased nitrogen availability to changes in other resources and combinations of resources (e.g., water, phosphorus)? We are aware of several multifactor resource manipulation experiments that have been recently initiated. These experiments can provide data to facilitate our analyses, but only if these efforts are coordinated properly. Finally, can competitive interactions (through manipulations of density) explain changes in diversity and community structure? Community-level competition experiments can directly test the effects of competition on diversity. Few direct tests have been conducted and results from these experiments indicate that current hypotheses to explain diversity-productivity relationships may not be supported.

Planning Grant Goals. Several unanswered questions concerning productivity-diversity relationships will provide a starting point for working group activities (Table 1). In order to address these questions, we have the following five objectives:

1) Involve a wider range of participants than currently are involved in the effort. We want to involve more people, particularly graduate students (many have approached us already interested in contributing), more sites/systems within and outside the LTER, and different types of resource manipulation experiments.

2) Develop standardized measures of functional (plant traits) and taxonomic (diversity) responses using the existing long-term data sets. These include patterns along natural productivity gradients and consequences of existing experimental manipulations of productivity. We plan to build upon our current dataset of nitrogen addition experiments, expanding to other resource manipulations and gradients with similar approaches.

3) Explore alternative measurements of diversity (evenness, diversity indices, richness), measures of community change (similarity indices, ordination scores), and productivity (NPP, resource availability, standing crop). Develop and test alternative analytical and multivariate modeling approaches. With the multi-year and multi-site database, incorporate temporal and spatial scales.

4) Based on the synthesis of existing LTER datasets, develop predictions about how the complement of plant traits changes over time in response to altered productivity (both increases and decreases), dynamics at different spatial and temporal scales, the mechanisms involved, and how these changes relate to diversity. Use analytical modeling approaches integrate theory with these empirical predictions.

5) Design coordinated, multi-site, standardized experiments to test these hypotheses. We will conduct experiments to generalize beyond studies of nitrogen addition to other limiting resources, both singly and in combination, as well as to competitive interactions and other mechanistic processes. We will also design experiments and monitoring to better link ecological feedbacks in human health and economics to the dynamics of plant response. Coordination and planning through a cross-site initiative will ensure comparable, standardized, and mechanistic studies.

Due to our working group’s history of collaboration, we expect that given support of the LTER network we can make great strides towards [g4] synthesis of existing data, development of new hypotheses and models, and testing those ideas. By utilizing the expertise available among the scientists in the LTER network, we were able to put together our current dataset and have submitted a research collaboration network (RCN) proposal to NSF (under review). Within the next 2 years, we expect to move onto the next crucial stage of large-scale synthesis and the implementation of cross-site experiments.

References Cited

Abrams, P.A. 1995. Monotonic or unimodal diversity productivity gradients- what does competition theory predict? Ecology 76: 2019-2027

Baer S.G., Blair JM, Collins SL, et al. 2003. Soil resources regulate productivity and diversity in newly established tallgrass prairie. Ecology 84: 724-735.

Davis, et al. 2000. Fluctuating resources in plant communities: a general theory of invisibility. J. Ecology 88: 528-534.

Gough, L., C.W. Osenberg, K.L. Gross, S.L. Collins. Fertilization effects on species density and primary productivity in herbaceous plant communities. Oikos 89(3): 428-439.

Gross, K.L., M.R. Willig, L. Gough, R. Inouye, S.B. Cox. 2000. Patterns of species density and productivity at different spatial scales in herbaceous plant communities. Oikos 89(3):417-427.

Knapp, A.K., P.A. Fay, J.N. Blair, et al. 2002. Rainfall variability, carbon cycling and plant species diversity in a mesic grassland. Science 298(5601): 2202-2205.

Mitttelbach G.G., C.F. Steiner, S.M. Scheiner, K.L. Gross, H.L. Reynolds, R.B. Waide, M.R. Willig, S.I. Dodson, L.Gough. 2001. What is the observed relationship between species richness and productivity? Ecology 82(9): 2381-2396.

Nemani RR, Keeling CD, Hashimoto H, et al. 2003. Climate-driven increases in global terrestrial net primary production from 1982 to 1999. Science 300: 1560-1563.

Oksanen, J., T. Tonteri. 1995. Rates of compositional turnover along gradients and total gradient length. J. of Vegetation Science 6(6): 815-824.

Stevens, M.H.H., W.P. Carson. 1999. Plant density determines species richness along an experimental fertility gradient. Ecology 80(2): 455-465.

Townsend, A., R. Horwarth, F. Bazzaz, et al. 2003. Human heath effects of a changing global nitrogen cycle. Frontiers in Ecology and the Environment 1: 240-246.

Vitousek, P. and R. Horwarth. 1991. Nitorgen limitation of land and sea: how can it occur? Biogeochemistry 13: 87-115.

Vitousek, P., J. Aber, R. Horwarth, et al. 1997. Human alteration of the global nitrogen cycle: causes and consequences. Ecological Applications 7: 737-750.

Waide, R.B., Willig, M.R., C.F. Steiner, G. Mittelbach, L. Gough, S.I. Dodson, G.P. Juday, R. Parmenter. 1999. The relationship between productivity and species richness. Annual Review of Ecology and Systematics 30:257-300.

Wolfe, A. and J. Patz. Reactive nitrogen and human health: acute and long-term implications. Ambio 31: 120-125.