At the leaf level, our findings indicated how the magnitude of the
Tr of the C
3 shrub
R. soongorica and the C
4 shrub
S. passerina changed within a day (
Figure 3). This does not agree well with Su et al. [
32], whose study suggested that the C
4 shrub
S. passerina had advantages over the C
3 shrub
R. soongorica through lower
Tr, especially when moisture is limited. The potential causes of this are the differences in plant growth status and environmental factors (e.g., meteorological and edaphic factors). Moreover, transpiration was lower in the C
4 than in the C
3 plants for a given photosynthetic capacity [
40]. Nevertheless, the results were in line with previous finding by Su et al. [
32], that the C
4 shrub
S. passerina has a higher WUE (
Figure 5B), since the C
4 pathway results in a higher WUE, with a higher efficiency utilization of low intercellular CO
2 concentrations at the leaf level [
39]. Stomata control determines the WUE of a plant by optimizing the water lost against the carbon gained [
41]. Su et al. [
32] also showed that the associated growth of the C
3 shrub
R. soongorica and the C
4 shrub
S. passerina decreased the
Tr of the whole community, whereas both
R. soongorica and
S. passerina existed as individuals in our study. C
3 and C
4 plants have different environmental adaptation strategies, responding differently to environmental factors [
42]. C
4 plants had an advantage compared to C
3 plants due to a decrease in transpiration under the same drought conditions [
43]. Furthermore, the C
4 photosynthetic mechanism equips plants with a valuable competitive advantage under conditions of high-light intensity, high temperatures and low water availability, exerting a tighter control over water balance [
44]. The ‘short-term’ stomata control responses of plants include stomata aperture changes in response to the availability of water, light, temperature, wind speed, and carbon dioxide [
40]. According to the correlations between
Tr and g
s and environmental factors, we can easily find that the
Tr of the C
3 shrub
R. soongorica and C
4 shrub
S. passerina was mainly affected by nonstomatal factors in May and September, respectively. It can be concluded that under the conditions of relatively adequate soil available water, the main influencing factors on the
Tr of the C
3 shrub
R. soongorica and the C
4 shrub
S. passerina are nonstomatal factors, i.e., environmental factors. The high temperature response observed in the
Tr of the C
4 shrub
S. passerina was clearly advantageous for the maintenance of high productivity in the desert, whereas the negative or slightly positive high temperature response observed in the
Tr of the C
3 shrub
R. soongorica was essential for its survival under harsh conditions (
Table 2 and
Table 3). When the air temperatures were below 35 °C, the average leaf temperatures were above the air temperature by an amount dependent on wind velocity, and the increasing wind diminished transpiration. When air temperatures were above 35 °C, the leaf temperatures were below the air temperatures, and increasing wind markedly increased transpiration [
45]. The negative response of the
Tr of the C
3 shrub
R. soongorica to the leaf-to-air vapor pressure deficit in June and July could be explained as an adaptation of the C
3 shrub
R. soongorica to long dry periods, during which the shrub was forced to maintain a minimum g
s to continue functioning [
21]. The negative response of the
Tr of the C
3 shrub
R. soongorica to the net carbon assimilation rate (A) was due to the high evaporative demand as well as the limitation of A by the intercellular CO
2 increases with temperature (
Table 3).
At the canopy level, both the
Tr of the C
3 shrub
R. soongorica and the C
4 shrub
S. passerina were well associated with the R
n and T
air, and the
Tr of the C
4 shrub
S. passerina is also sensitive to changes in VPD (
Table 4). Transpiration is controlled by radiation as the rate of production of water vapor inside the leaf is driven by absorbed radiation [
46]. Furthermore, rainfall was the main factor influencing the transpiration amount of the C
3 shrub
R. soongorica and the C
4 shrub
S. passerina during the growing season. Canopy height had a significant effect on the transpiration of the C
3 shrub
R. soongorica, and the transpiration of the C
3 shrub
R. soongorica was influenced by canopy height, canopy projection area and new shoot length (
Table 5). As for edaphic influences on the transpiration of the C
3 shrub
R. soongorica and the C
4 shrub
S. passerina during the growing season, our study observed the soil matric potential and soil temperature impact on transpiration. The results indicated that the transpiration responses of the C
3 shrub
R. soongorica and the C
4 shrub
S. passerina to soil matric potential were opposite, since the C
3 shrub
R. soongorica had less plant-available soil water compared to the C
4 shrub
S. passerina (
Figure 2).Transpiration was not affected by soil moisture under well-watered conditions [
47]. When the soil matric potential exceeds the field capacity (e.g., pF > 2.5), plants start reducing transpiration by closing their stomata to prevent internal water loss [
48]. In summary, how the transpiration of xerophytic shrubs responds to soil moisture depends on the available soil water content. The transpiration of the C
3 shrub
R. soongorica and the C
4 shrub
S. passerina responded positively to soil temperature, and this response was statistically significant for
S. passerina (
Table 6). A possible explanation for this may be that the soil water content was negatively correlated with soil temperature at moderate-to-high water contents [
49]. Transpiration increases with high soil temperature due to the fact that higher temperature promotes the absorption of soil water by the roots of the plant [
50].
The purpose of our study was to demonstrate whether it is feasible to approximately estimate the transpiration of xerophytic shrubs using the scaling approach in the absence of a lysimeter. The results proved successful for up-scaling the transpiration of C
3 and C
4 xerophytic shrubs from the leaf to the canopy level using the unified scaling factor (e.g., leaf area).The results are comparable to Huang et al. [
25], Zhao and Zhao [
13], and Gitz III et al. [
13], whose research involved up-scaling from the leaf to the individual or canopy level. Although long-term transpiration continuously determined by a LI-6400XT portable photosynthesis system is difficult due to the influence of micrometeorological factors and artificial operation, there is no doubt that transpiration can be effectively and accurately measured using the LI-6400XT method within a short time. This may be explained by the high coefficient of determination between the canopy transpiration rate found using the LI-6400XT and that found by lysimeter measurement (R
2 = 0.7314 for
R. soongorica and R
2 = 0.5892 for
S. passerina) (
Figure 6), between the daily transpiration rate using the LI-6400XT and that found by using stem heat balance measurement (R
2 = 0.67 for
C. korshinskii and R
2 = 0.77 for
A. ordosica) [
8], between canopy transpiration found using LI-6400XT and that found using the FAO-56 dual crop coefficient method (R
2 = 0.66 for maize) [
13], as well as between the canopy transpiration found using the LI-6400XT and that found using the canopy evapo-transpiration and assimilation (CETA) chamber system (R
2 = 0.94 for cotton) [
17]. Transpiration up-scaling from leaf-level measurements was much higher than lysimeter measurement due to many errors. For example, a certain error was due to the selected leaves where the leaf transpiration was measured. The leaf area obtained from photograph processing and integration also led to some error. Moreover, the monitored leaf transpiration values often fluctuate during measurement, which also leads to some errors. This discrepancy may also be due to the error caused by the fact that scaling up leaf-level measurements assumes that all leaves within the canopy receive equal ambient radiation [
51].