TEMPORAL ASYNCHRONY IN CANOPY-CLIMATE RELATIONSHIP
Growing season length is highly dependent on the availability of water for photosynthesis, when soil and air water content increase dramatically at the end of the dry season, the forest vegetation responds by putting out leaf buds and if the conditions remain moist the leaves expand rapidly into a mature canopy. When plotting the changing canopy greenness response against the meteorological variables at the start of the growing season the relationship is relatively weak showing Pearson correlation coefficients between 0.2 and 0.5 (figure R2). As we shift the synchrony of these climate-canopy measurements though we begin to see an increasingly strong relationship between the two systems whereby, after about 10 days the correlation between soil moisture in Mexico and all four meteorological variables in Brazil nearly doubles (Figure R2. This temporal asynchrony between the foliage response and water availability is essentially a physiological time lag in how fast the plants can change morphology through leaf expression to utilize the water for photosynthesis, growth, and reproduction.
Once leaf development is complete and the canopy is fully mature the lag-time in response to atmospheric conditions is essentially removed as the plant's stomata are now energetically and chemically coupled to their environment when they are open for leaf level photosynthesis. The rate of photosynthesis is dependent upon the stomatal conductance which increases as a function of vapor pressure gradient between the leaf and the surrounding air space (Sack and Scoffoni, 2012). As such we see a much greater relationship between canopy greenness and meteorological conditions during the third phenophase with a fully developed canopy. The changes in canopy greenness during the peak of the growing season is likely not a function of leaf area increase as we see during the green-up period but rather a change in leaf angle due to water stress (Chaves et al, 2002; Oosterhuis et al, 1985). When a plant's water budget is stressed the mean leaf angle increases to a more vertical orientation, this is a common adaptation in semi-arid regions where solar radiative forcing is high and light interception is greatly reduced when leaf angle increases from wilting. A change in leaf angle distribution in a plant canopy will affect light interception and reflectance and may explain the decreasing NDVI values with increasing temperature, vapor pressure, or decreasing relative humidity seen during the growing season in figure R3. It is worth noting that while temperature does not appear to influence canopy greenness during leaf maturity at the coastal TDF, it does affect the continental dry forest canopy where humidity levels are less stable and water stress is an insurmountable physiological barrier at higher temperatures (figure R3).
As the rainy season ends and water resources become scarce, both ground water and water vapor, the canopy becomes increasingly stressed and unable to afford the physiological costs of having leaves which tend to lose water when atmospheric levels are low. We can see from the soil moisture and relative humidity scatterplots in figure R1 that daily average moisture levels decrease significantly before a significant reduction of canopy greenness is observed. Then as water availability is at its lowest the canopy greenness rapidly decreases showing little correlation with the soil water content. This indicates another temporal asynchrony between environmental conditions and plant physiology but for different reasons than we have during the leaf flushing period. The plants wan to maintain leaves as long as possible to utilize the sunlight for creating energy reserves and they even maintain them well after environmental water resources are scarce. It is possible that going into the dry season the trees that have deep root access to ground water are tapping that resource to maintain leaves or water stored in the plant may also be utilized. Either way, after 2 months of no rainfall we see the onset of leaf senescence for both sites, after this time frame the water stress must exceed the resources the plants have to continue productive photosynthesis. Leaf drop is a long process as the canopy greenness can be seen to decrease over a 2 to 3 month period before leafless dormancy is fully observed. This may be due to the different water stress tolerance of the tree species in the canopy as some will lose their leaves before others.
Once leaf development is complete and the canopy is fully mature the lag-time in response to atmospheric conditions is essentially removed as the plant's stomata are now energetically and chemically coupled to their environment when they are open for leaf level photosynthesis. The rate of photosynthesis is dependent upon the stomatal conductance which increases as a function of vapor pressure gradient between the leaf and the surrounding air space (Sack and Scoffoni, 2012). As such we see a much greater relationship between canopy greenness and meteorological conditions during the third phenophase with a fully developed canopy. The changes in canopy greenness during the peak of the growing season is likely not a function of leaf area increase as we see during the green-up period but rather a change in leaf angle due to water stress (Chaves et al, 2002; Oosterhuis et al, 1985). When a plant's water budget is stressed the mean leaf angle increases to a more vertical orientation, this is a common adaptation in semi-arid regions where solar radiative forcing is high and light interception is greatly reduced when leaf angle increases from wilting. A change in leaf angle distribution in a plant canopy will affect light interception and reflectance and may explain the decreasing NDVI values with increasing temperature, vapor pressure, or decreasing relative humidity seen during the growing season in figure R3. It is worth noting that while temperature does not appear to influence canopy greenness during leaf maturity at the coastal TDF, it does affect the continental dry forest canopy where humidity levels are less stable and water stress is an insurmountable physiological barrier at higher temperatures (figure R3).
As the rainy season ends and water resources become scarce, both ground water and water vapor, the canopy becomes increasingly stressed and unable to afford the physiological costs of having leaves which tend to lose water when atmospheric levels are low. We can see from the soil moisture and relative humidity scatterplots in figure R1 that daily average moisture levels decrease significantly before a significant reduction of canopy greenness is observed. Then as water availability is at its lowest the canopy greenness rapidly decreases showing little correlation with the soil water content. This indicates another temporal asynchrony between environmental conditions and plant physiology but for different reasons than we have during the leaf flushing period. The plants wan to maintain leaves as long as possible to utilize the sunlight for creating energy reserves and they even maintain them well after environmental water resources are scarce. It is possible that going into the dry season the trees that have deep root access to ground water are tapping that resource to maintain leaves or water stored in the plant may also be utilized. Either way, after 2 months of no rainfall we see the onset of leaf senescence for both sites, after this time frame the water stress must exceed the resources the plants have to continue productive photosynthesis. Leaf drop is a long process as the canopy greenness can be seen to decrease over a 2 to 3 month period before leafless dormancy is fully observed. This may be due to the different water stress tolerance of the tree species in the canopy as some will lose their leaves before others.
PRECIPITATION AND TROPICAL SEMI-ARID BIOMES
Figure D1: Biomes by mean annual temperature and precipitation as observed by NASA's Moderate
Resolution Imaging Spectro-radiometer and validated by ground observations at FluxNet sites. Seasonally Dry Tropical Forests are seen to be in the same temperature zone as tropical rain forest and savannah biomes but with an intermediate rate of precipitation. Source: FluxNet, 2012.
Figure D1 demonstrates the relationship between mean annual temperature and precipitation across the globe, where each remotely sensed 10x10km land surface MODIS pixel is colored by biome classification. FlexNet is a network of ground based satellite validation sites pictured in black. Aboveground biomass increases with both temperature and precipitation across biomes. Looking at the upper range on the temperature scale, where mean annul temperatures exceed 17°C, we see the largest spread in annual precipitation rates from 0mm to nearly 5m. Tropical seasonal forests are situated in a disctinct rainfall regime bewteen 1500mm and 2500mm, above which the highly productive tropical rain forests exist and below this we see less productive tropical savannahs and shrublands. The division betwen these biomes is depicted here to be different only by a couple hundred millimeters less rainfall a year on average. Since tropical seasonal forests exist in a balance between being too dry to support rainforest plant communites but too wet to have savannah vegetation. We might expect that with a warmer climate changing precipitation patterns across the tropics, regions with seasonlly dry forests could be forced into an ecological transition between biomes. Currently, there is no research describing a threshold for inducing such a bioclimatic transition in tropical dry forests, as such we need to better understand potential mechanisms underlying semi-arid biome productivity as a function of whole ecosystem responses to climate change.
Figure D2: Precipitation Projections with Climate Change. Global prediction for changing rainfall rates by the end of the 21st century. This is Climate Change model (CM2.1) is produced by the Geophysical Fluid Dynamics Laboratory as a scenario of minimal reduction in current CO2 emmisions (2005). The CM2.x model series is a coupled ocean-atmospheric general circulation model and was one of the leading climate models used by the Intergovernmental Panel on Climate Changes fourth assessment (Journal of Climate, 2006). Most of the tropical dry forest distributions are found where the model predicts a decrease in precipitation, specifically the southeastern region of Brazil.
CONCLUSIONS
Differences in canopy state and pheno phase transition characteristics in relation to local climate can be seen clearly when we look at the time-series trends (Figure D2). While both monitoring sites are semi-arid tropical environments, the coastal Mexican dry forest experiences little fluctuation in air temperature, relative humidity, and vapor pressure throughout the year as mediated by the ocean proximity. In contrast, the atmospheric conditions of the continental Brazilian dry forest change drastically through the year in concert with rainfall patterns and transient humid air front movements across the south american interior. As a result the plant physiology and leaf phenology of the two sites differ with respect to changes in local meteorology. We have seen that the Brazilian dry forest canopy shares a stronger relationship with both ground and air water availability while the vegetative phenology of the Mexican dry forest is really only driven by rainfall events and soil moisture content. The greater amount of precipitation, by about 100mm, at the Brazilian site during this given year appears to also have allowed for a longer growing season. So while phenological phase transition biometeorology of these two tropical dry forest sites tend to behave differently, the annual productivity of the ecosystem is very much driven by the frequency and extent of rainfall input during the wet season.
Changes in terrestrial precipitation patterns are expected across the Americas as the world gets warmer, and with this the productivity of semi-arid ecosystems will respond as a function of water availability. With most of the Latin American population in semi-arid regions where tropical dry forests exist, it is important to track the productivity of these environments to help anticipate threats to food and water security. Through the use of remote sensing and ground based monitoring we can better understand the limitations and long term capacity of dry forest ecosystem resilience to climate change in order to better plan forest resource management initiatives for the future.
Changes in terrestrial precipitation patterns are expected across the Americas as the world gets warmer, and with this the productivity of semi-arid ecosystems will respond as a function of water availability. With most of the Latin American population in semi-arid regions where tropical dry forests exist, it is important to track the productivity of these environments to help anticipate threats to food and water security. Through the use of remote sensing and ground based monitoring we can better understand the limitations and long term capacity of dry forest ecosystem resilience to climate change in order to better plan forest resource management initiatives for the future.
REFERENCES
Chaves et al. 2002.How Plants Cope with Water Stress in the Field? Photosynthesis and Growth. Annals of Botany, 89(7): 907-916.
Oosterhuis, DM, S Walker, and J Eastham. 1985. Soybean Leaflet Movements as an Indicator of Crop Water Stress. Crop Science, 6: 1101-1106.
Sack, L. and C. Scoffoni. 2012. Measurement of leaf hydraulic conductance and stomatal conductance and their responses to irradiance and dehydration using the Evaporative Flux Method (EFM). Journal of Visualized Experiments, (70): 4179.
NOAA Geophysical Fluid Dynamics Laboratory: http://nomads.gfdl.noaa.gov/CM2.X/
Oosterhuis, DM, S Walker, and J Eastham. 1985. Soybean Leaflet Movements as an Indicator of Crop Water Stress. Crop Science, 6: 1101-1106.
Sack, L. and C. Scoffoni. 2012. Measurement of leaf hydraulic conductance and stomatal conductance and their responses to irradiance and dehydration using the Evaporative Flux Method (EFM). Journal of Visualized Experiments, (70): 4179.
NOAA Geophysical Fluid Dynamics Laboratory: http://nomads.gfdl.noaa.gov/CM2.X/