Carbon (C) and nitrogen (N) are major components of all plant tissues, and their metabolism is intimately linked to plant life function (Griffiths, 1998). Stable C and N isotopic signatures of ecosystem has helped elucidate the pathways and rates of nutrient exchange about biotic and abiotic controls on the cycling of C and N (Reviewed by Dawson et al., 2002). The distinction between δ13C values of C3 and C4 plants is due to different photosynthetic pathways to fix atmospheric CO2 ([Farquhar et al., 1982], [O'Leary, 1995] and [Vogel, 1980]). C3 plant δ13C values range globally from − 22.0 to − 37.0‰, averaging around − 27.0‰, while C4 plants have δ13C values between − 9.0 and − 15.0‰, with a mean of − 12.5‰ ([Farquhar et al., 1982], [O'Leary, 1988] and [Vogel, 1980]). δ13C values of C3 plants also vary between individuals, species, and populations, because environmental factors and different physiological adaptations and responses lead to further isotopic discrimination during photosynthesis, mainly by affecting CO2 conductance (through the stomata) and carboxylation rates ([Farquhar et al., 1982], [Farquhar et al., 1989], [O'Leary, 1988], [O'Leary, 1995] and [Tieszen, 1991]). Environmental factors, such as solar radiation levels, temperature, water availability, and water-use efficiency, are widely quoted as the principal factors underlying carbon isotopic variations within C3 plants ([Farquhar et al., 1988], [Farquhar et al., 1989], [Heaton, 1999], [van der Merwe and Medina, 1991], [O'Leary, 1995] and [Tieszen, 1991]). Similar environmentally-induced variations in C4 plants are less pronounced, because CO2 is fixed early during the photosynthetic process, limiting further fractionation or discrimination ([O'Leary, 1988] and [Vogel, 1980]). However, C4 plants following alternative photosynthetic sub-pathways discriminate differently against 13C, resulting in small (− 1‰) but significant variations in δ13C between C4 sub-types ([Hattersley, 1982] and [Sage et al., 1999]).

Plants will not respond to environmental change as communities; rather, each species will respond independently according to its sensitivity to disturbance regimes or variations in moisture (Spear et al., 1994). The actual sensitivity of response varies significantly among locations and /or species. Thus, the species-specific responses may affect the isotopic signatures and dampen or confound the effects of drought on C isotopic discrimination ([Schulze et al., 1998] and [Schulze et al., 1999]). Moreover, life form, reflecting different strategies to cope with xeric environment, had a significant fraction of the variation in plant carbon isotope composition ([Brooks et al., 1997] and [Chen et al., 2005]). However, for C4 plants, the δ13C values are not sensitive to water availability (Farquhar et al., 1982) and the correlation between the plant δ13C and water availability (e.g. precipitation) commonly obtained for C4 plants is usually absent ([Schulze et al., 1996] and [Swap et al., 2004]).

Nitrogen isotope ratios (δ15N) provide information related to nitrogen cycling within ecosystems ([Högberg and Alexander, 1995], [Roggy et al., 1999] and [Ometto et al., 2006]). Plant N isotopic discrimination is also related to the availability of nutrients and water availability ([Tilman, 1988], [Swap et al., 2004] and [Liu et al., 2007]) and indicative of N cycling on different spatial and temporal scales (Nadelhoffer and Fry, 1994). An enrichment of δ15N signatures in plant samples associated with arid regions has been demonstrated for sites within the arid desert environments ([Handley et al., 1999] and [Swap et al., 2004]). This enrichment suggests different biogeophysical processing and cycling of N caused by decreased water availability. Moreover, Schulze et al. (1994) found that at Alaskan sites where nitrogen is limiting, different life forms (or genus) exploited different nitrogen pools within the ecosystem.

The ecosystem cycling of nitrogen (N) and carbon (C) are so interlinked that changes in the availability of one is likely to affect and be affected by the availability of the other (BassiriRad et al., 2003). Plant δ13C and δ15N signatures were regulated by some concurrent factors. For example, photosynthesis was correlated with water availability, poor water availability would decrease the ability of photosynthesis (C assimilation), further result in reduction of productivity and N requirement or assimilation, in which different δ13C and δ15N variations occurred (Yoneyama et al., 2001). The positive relationship found in an aridity gradient in Namibia (Schulze et al., 1991), and this relationship may be better interpreted as a positive correlation between water supply and amount of N2-fixation (Handley and Raven, 1992 L.L. Handley and J.A. Raven, The use of natural abundance of nitrogen isotopes in plant physiology and ecology, Plant Cell Environ. 15 (1992), pp. 965–985. Full Text via CrossRef | View Record in Scopus | Cited By in Scopus (227)Handley and Raven, 1992). Schulze et al. (1991) interpreted a highly significant (but strongly scattered) trend between δ15N and δ13C as decreased water use efficiency in African N2-fixing trees (as opposed to non-fixers) which they thought was possibly caused by the extra cost of supplying carbohydrates for diazotrophy. However, to date, there was little study about the artificial sand-binding ecosystems.

The planting of sand-binding vegetation in Shapotou region at the south-eastern edge of the Tengger Desert began in 1956. Over the past 50 years, it has not only insured the smooth operation of the Baotou–Lanzhou railway in the sand dune section but also played an important role in the restoration of the local ecosystems; therefore, it is viewed as a successful model of controlling desertification and the ecological restoration along the transport line in the arid desert region of China ([Xiao et al., 2003] and [Li et al., 2006]). After the long-term succession, the mobile dunes change towards a landscape of the steppe desert with increase of biological processes and accumulation of soil organic matter under the extremely conditions (Xiao et al., 2003). There is a new ecohydrological processes of soil–plant system due to transpiration of revegetation ([Xiao et al., 2004] and [Li et al., 2004]), which created favorable conditions for colonization by many herbaceous (Xiao et al., 2003). The colonization of annual and perennial plant species promoted the succession and restoration of the vegetation towards herb-dominant vegetation ([Xiao et al., 2004] and [Li et al., 2004]). Thus, in this extreme arid region, where water stress is one of the key limiting factors of the ecosystem development, and the contradiction of supply and demand in soil water and nutrients resource is prominent increasingly. Recently, the plant water-use efficiency, plants nutrients and their connections were explored in these artificial sand-fixing vegetation by measuring foliar δ13C and nitrogen, phosphorus and potassium nutrition ([Zhao et al., 2007a] and [Zhao et al., 2007b]), which mainly focus on the dominant shrub plants. As yet, the general differences of carbon and nitrogen isotope ratios of plant species among the different aged-fixed districts (C3 and C4 plants) are scare still.

The existing sand-fixing vegetation zones establishing in different years (1956, 1964, 1981, and 1987), the transitional zone between revegetation and shifting sand dunes, and the adjacent natural vegetation with heterogeneity of soil moisture and nutrition provide conditions to examine the patterns in δ13C and δ15N distributions as it relates to C3 and C4 pathways. Furthermore, the site- and species-specific water-use pattern of different photosynthetic pathway plants by isotopic means can give implications which species could hold a competitive advantage in terms of WUE and nutrient use efficiency (NUE) in a new balance of soil water and nutrition in the current succession stage.

The present study examines the δ13C and δ15N natural abundance of the coexistent species (C3 and C4) to determine: (1) the basal information of the dual isotopic signatures of C3 and C4 plants in different aged artificial sand-binding vegetation, transitional and the natural microhabitats; (2) effects of life form, microhabitation and bio-crust on isotopic variations; (3) the relationships between δ13C and δ15N values of two photosynthetic pathways. Further, the species-specific patterns of both isotopic signatures were investigated. The results can give the insights into ecosystem processes and provide the base information for biochemical cycles in arid desert regions.

Carbon isotope ratio (δ13C) and nitrogen isotope ratio (δ15N) of leaf in shrubs and overground matter in herbage were measured on plant species occurring in different aged artificial sand-binding microhabitats, as well as in natural habitat at the south-eastern margin of the Tengger Desert, China. Both δ13C and δ15N of C3 and C4 plants varied widely (− 28.12 ≤ δ13C (‰) ≤ − 23.77 and − 4.45 ≤ δ15N (‰) ≤ 3.66 for C3 species, respectively; and − 15.79 ≤ δ13C (‰) ≤ − 12.63 and − 7.56 ≤ δ15N (‰) ≤ 1.08 for C4 species, respectively), representing the different photosynthetic pathway (C3/C4) environmental controls. The relative abundance of C4 species increased over the development of the sand-fixing vegetation, and the significant differences among sites in δ15N of C4 plants were found. Among the microhabitats, the isotopic pattern of artificial sand-fixing community is a cluster compared to that of natural vegetation and transitional zones in C4 plants; however, this pattern of C3 plants is different significantly. The biological soil crust had significant effect on δ13C of C3 plants, not for δ13C of C4 plants and δ15N of both C3 and C4 species. Our results also describe a distinct pattern of dual-isotopic signatures, which indicated that a different water-use source and soil nitrogen compartment occurred and may promote the coexistence of different life forms in extreme poor water and nutrients ecosystems. With respect to C4 plants, the gramineous plants based on family — level exhibit obviously negative isotopic values (both δ15N and δ13C) than those of chenopodiaceous plants, indicating a high water-use efficiency of gramineous C4 plants and more 15N depleted, and these differences may be a result of different functional groups of C4 species. For C3 plants, the chenopodiaceous species also show more 15N enrichment than those of plants from other families (e.g. composite, leguminous plants and others). Contrast to previous studies, the δ15N values of leguminous plants were not closer to the δ15N of atmospheric nitrogen gas, indicating that leguminous plants in research region used nitrogen mainly from other N sources such as soil, rain/dust, not mainly from biological fixed nitrogen. The correlation between δ13C and δ15N of C3 plants (except leguminous plants) were negative significantly (P = 0.001), while there were positive relationship between δ13C and δ15N in C3 leguminous plants (P = 0.003) and C4 plants (P < 0.0001), indicating that in the water and nutrient limiting region the carbon and nitrogen discriminations are dependent and linked intimately, and vary with the plant species and photosynthetic pathway.