Legacy of historic ozone exposure on plant community and food web structure

Abstract

Information on whole community responses is needed to predict direction and magnitude of changes in plant and animal abundance under global changes. This study quantifies the effect of past ozone exposure on a weed community structure and arthropod colonization. We used the soil seed bank resulting from a long-term ozone exposure to reestablish the plant community under a new low-pollution environment. Two separate experiments using the same original soil seed bank were conducted. Plant and arthropod richness and species abundance was assessed during two years. We predicted that exposure to episodic high concentrations of ozone during a series of growing cycles would result in plant assemblies with lower diversity (lower species richness and higher dominance), due to an increase in dominance of the stress tolerant species and the elimination of the ozone-sensitive species. As a consequence, arthropod-plant interactions would also be changed. Species richness of the recruited plant communities from different exposure histories was similar (≈ 15). However, the relative abundance of the dominant species varied according to history of exposure, with two annual species dominating ozone enriched plots (90 ppb: Spergula arvensis, and 120 ppb: Calandrinia ciliata). Being consistent both years, the proportion of carnivore species was significantly higher in plots with history of higher ozone concentration (≈3.4 and ≈7.7 fold higher in 90 ppb and 120 ppb plots, respectively). Our study provides evidence that, past history of pollution might be as relevant as management practices in structuring agroecosystems, since we show that an increase in tropospheric ozone may influence biotic communities even years after the exposure.

Fig 1. Relative abundance of Spergula arvensis, Calandrinia ciliata and other species in plant communities selected under different episodic concentrations of tropospheric ozone.

Ozone concentrations in the open top chambers during long-term exposure were 0 ppb (white bars), 90 ppb (grey bars) and 120 ppb (dark bars). Data represents species abundances during the first (a) and second (b) year of experiment after original soil seed bank was transplanted to a common natural field environment. Relative abundance for each plant species was calculated as the summed abundances of each plant species for a particular year/total number of seedlings recorded in the plot (n = 3). Error bars represent standard error. Year a ANOVA P _species < 0.01, P_ozone 0.016, P_{species x ozone} 0.034; year b ANOVA P _species < 0.01, P_ozone 0.024, P_{species x ozone} 0.042

Fig 2. Relative abundance of arthropods corresponding to different functional groups colonizing field plots originated from plant communities selected under different episodic concentrations of tropospheric ozone.

Ozone concentrations in the open top chambers during long-term exposure were 0 ppb (white bars), 90 ppb (grey bars) and 120 ppb (dark bars). Data represents arthropod abundances during the first (a) and second (b) year of experiment, after original soil seed bank was transplanted to a common natural field environment. Arthropods were assigned to different groups (guilds) according to larval feeding strategy. Relative abundance for each guild was calculated as the summed abundances of insects of that guild for a particular year/total number of insects recorded in the plot (n = 3). Error bars represent standard error. Year a ANOVA P _guild < 0.01, P_ozone 0.03, P_{guild x ozone} 0.021; year b ANOVA P _guild 0.02, P_ozone 0.014, P_{guild x ozone} 0.07

Fig 3. Relationship between the arthropod diversity (Shannon index H´) and the number of plant species.

Plant species were recorded in the field plots where soil seed bank selected under different episodic ozone concentrations [0 (white circle), 90 (grey square) and 120 (black triangle) ppb] was sown. Arthropods naturally colonized the plots. Each data point corresponds to one plot coming from a particular ozone concentration in one of two years of experiment. Spearman´s correlation 0 ppb r_s = 0.71, n = 6, P < 0.01; 90 ppb r_s = 0.02, n = 6, P = 0.75; 120 ppb, r_s = 0.10, n = 6, P = 0.68