SUPPORTING INFORMATION Phosphorus accumulates faster than nitrogen globally in freshwater ecosystems under anthropogenic impacts Zhengbing Yan1,2, Wenxuan Han2*, Josep Peñuelas3,4, Jordi Sardans3,4, James J. Elser5, Enzai Du6, Peter B. Reich7,8, Jingyun Fang1 1 Department of Ecology, College of Urban and Environmental Sciences, Peking University, Beijing 100871, China 2 College of Resources and Environmental Sciences, China Agricultural University, Beijing 100193, China 3 CSIC, Global Ecology Unit CREAF-CSIC-UAB, Cerdanyola del Vallès, 08193 Barcelona, Catalonia, Spain, 4 CREAF, Cerdanyola del Vallès, 08193 Barcelona, Catalonia, Spain 5 School of Life Sciences, Arizona State University, Tempe, AZ 85287, USA 6 College of Resources Science & Technology, and State Key Laboratory of Earth Surface Processes and Resource Ecology, Beijing Normal University, Beijing 100875, China 7 Department of Forest Resources, University of Minnesota, St. Paul, MN 55108, USA 8 Hawkesbury Institute for the Environment, Western Sydney University, Penrith, 2751 New South Wales, Australia 1
To whom correspondence should be addressed; E-mail:
[email protected]
This file includes: Supplementary discussion Tables S1–S10 Figures S1-S10 Appendix Data sets
1
Supplementary discussion Effects of functional group on N and P stoichiometry in freshwater macrophytes ANOVA results showed that there were significant differences in plant N and P stoichiometry across functional groups (Table S3). Statistical comparisons between functional groups confirmed a relatively constrained range of plant N:P ratios compared to N and P concentrations (Table S3). Similar to their terrestrial cousins (Güsewell 2004), freshwater graminoids had lower P concentration and higher N:P than forbs. Emergent plants often had more supporting tissue with less nutrient concentration (Fernández-Aláez et al. 1999; Güsewell & Koerselman 2002), which might cause the lower tissue N and P concentrations than plants of other life forms. Freely floating plants had rapid growth rate and were often dominant in the eutrophic water bodies (Bornette & Puijalon 2011). Growth Rate Hypothesis states that organisms with rapid growth rate tend to have high tissue N and P concentration (Sterner & Elser 2002). Freely floating plants should thus have especially high N and P concentrations than plants of the other lifeforms, and might be suitable for purification of waste waters given that floating plants can easily be separated and harvested from waterbodies (Dykyjová1979). Moreover, higher N and P concentrations occurred in ferns than in seed plants (in opposition to the pattern in terrestrial plants, Han et al. 2005), and in dicotyledons than in Monocotyledon, suggesting that plant N and P stoichiometries varied among these phylogenic groups.
Latitudinal patterns of N and P stoichiometry in freshwater and its macrophytes. Latitudinal patterns of N and P stoichiometry in freshwater macrophytes were much weaker in contrast with previous proposals for terrestrial plants (Table S6; Fig. S8; Reich & Oleksyn 2004; Han et al. 2005 & 2011; Borer et al. 2013; Chen et al. 2013), which probably can be attributed to the diversity of habitat-specific biological, edaphic and anthropogenic influences on nutrient availabilities in freshwater ecosystems, and the relative similarity (less heterogeneity) of the habitat environments in water vs. land because of the general connectivity of water systems. Likewise, congeneric aquatic species generally show less geographic variation in many traits than their terrestrial cousins. Geographic gradients of species composition, climatic factors and soil nutrients jointly determine the latitudinal pattern of leaf N and P stoichiometry in terrestrial plants (Reich & Oleksyn 2004, Han et al. 2005 & 2011, 2
Chen et al. 2013). The latitude pattern of N and P stoichiometry in freshwater macrophytes, however, may mainly depend on geographic gradient of nutrient availability of freshwater water bodies (Downing & McCauley 1992). Water TN and TP concentrations in freshwater water bodies also decreased, and TN:TP ratios increased with increasing absolute latitude (Table S6; Fig. S8; Abell et al. 2012). The homeostasis of plant N was generally higher in contrast with plant P, presumably resulting in the non-significant relationship between plant N concentration and the absolute latitude (Sterner & Elser 2002; Han et al. 2011). Given that nutrient availability in freshwater ecosystems was easily altered by external condition, the derived latitudinal pattern might be caused in part by existing geographical gradient of humaninduced N and P inputs (Van Drecht et al. 2009, Potter et al. 2010). Moreover, the internal nutrient cycling of freshwater water bodies is quite different between tropics and temperate zone. N loss from water bodies in tropics is stronger than that in temperate zone via denitrification due to higher frequency of water de-oxygenation with higher temperatures, but P mobilization in tropics is more rapid (Downing et al. 1999), leading to lower water TN:TP ratio in tropics than in temperate zone.
The Decreasing fertilizer N:P use in China Indeed, the fertilizer N:P ratio showed a negative trend from 1985 to 2014, although there were slight differences among the three data sources and some fluctuations since 2007 (Fig. S9). There are three potential causes for the decreasing fertilizer use N:P ratio (Fig. S9). First, the consumption structure of fertilizers has shifted in recent decades, with an increase in cash crop areas but a decrease in grain crop areas in China, resulting in lower fertilizer N:P (Yang 2004); because most cash crops have higher total fertilizer use and require lower fertilizer N:P than grain crops (Yang 2004; Zhang & Zhang 2008). Second, land use changes induced by the ‘Conversion of Cropland to Forest and Grassland Program’ in 174 counties in western China led to the decline in N fertilizer use (Yang 2004). Third, agronomic management projects (i.e. balanced fertilization program) decreased the fertilizer use N:P from 4.35 to 2.94 in seven main crops in Hebei, Heilongjiang, Shanxi, Jiangsu, Zhejiang, Hunan and Sichuan provinces (Yang 2004).
3
Supplementary Tables S1-S3 Table S1 Summary of reduced major axis (RMA) regression results between P and N concentrations (i.e. log10 P=α*(log10 N)+β) of freshwater macrophytes for all raw data pooled. EA represented the Euro-America, which consisted of sites from Europe and the USA. n
αRMA (95% CI)
βRMA (95% CI)
r2
p
All
909
1.42 (1.34; 1.50)
-1.48 (-1.59; -1.37)
0.21
