Total plant volume was calculated as the volume of a pyramid height and the two canopy widths . Data on canopy dieback for each individual were collected in fall 2016. Dieback was assessed as the actual percent of “non-green” vegetation, defined as yellow, brown, and black/gray leaves, as well as bare/defoliated stems within the canopy. Percent dieback of each canopy was estimated by two-to-three researchers viewing multiple angles of each shrub, and final estimates were determined after thorough consultation. Site dieback was then calculated as the mean of all selected shrubs within a site. Entire stand dieback was also estimated using a combination of ground-level assessments and, when available, aerial drone photographs. These were used to confirm that canopy dieback of individual shrubs were collectively representative of whole stand dieback.Data were again collected approximately every 4-6 weeks throughout the summer dry season, this time from June through October. XPP measurements were taken using the same methods as described above. Additionally, we measured daytime gas exchange to gain a better understanding of plant function throughout the dry season and as water availability declined. Anet was measured using a Licor 6400XT . CO2 was set to 400ppm, and photosynthetically active radiation was set between 1400-1800, to reflect maximum photosynthesis conditions at peak daytime hours. All measurements were taken between 9am and 11am DST, typically on the morning after the predawn measurements were made.
A single fully-expanded leaf was chosen per plant, procona flower transport containers and two readings were recorded and averaged later during data processing. Leaves were traced in the field, and tracings were brought back to the lab and scanned to calculate leaf area within the chamber using ImageJ software . Gas exchange data were then adjusted by calculated leaf areas. Our FMS 2 system broke in August, and we were therefore unable to collect any Fv/Fm data after the July fieldwork. This study sought to understand spatial patterns of A. glauca canopy dieback across the landscape during an unprecedented drought, and track the progression of dieback and eventual mortality in this classically drought-tolerant shrub. Our data support the hypothesis that dieback is related to water stress from drought, and that this varied across the landscape, but there was considerable variation between sites and across time. While aspect was significantly correlated with dieback across all years, we found no consistent significant effect of elevation on dieback until 2019, late in drought period. Further, while dieback generally increased across all sites from 2016 to 2019, we observed no new incidences of mortality during this study, suggesting that in this landscape A. glauca individuals are resilient to this punctuated but overall prolonged drought.Consistent with our predictions, elevation and aspect appeared to both have significant correlations with XPP and Anet. While these correlations were weak, their significance lends support to our hypothesis that plant stress is related to these landscape variables. These relationships also provide evidence that elevation and southwestness may be used as tools for identifying areas where plant canopies are more vulnerable drought.
The low correlation coefficients may be due to extreme landscape heterogeneity in the study region including heterogeneity in substrate rockiness, soil accumulation and topographic concavity, confounded by a relatively small sample size within each site. There are many microclimate variables that were not included in this study and that may contribute to variations in water stress, including site temperature, vapor pressure deficit, and fog patterns. Additionally, though we did not find plant size to be correlated with drought stress or dieback, we believe this is due to our efforts to select even-aged individuals for monitoring, while previous studies have found significant effects of plant canopy diameter on dieback and mortality . Also, because A. glauca recruit from seed after fire, individuals within a stand tend to be of similar size and age. Therefore, we recommend that future studies include greater variation in stand age so that relationships between age/size class and dieback can be better evaluated.Elevation was not shown to be highly correlated with dieback in all years of the study, and there did not appear to be a correlation between dieback annual precipitation. Yet, some important landscape patterns were revealed. Our hypothesis that dieback severity is related to landscape variables associated with water availability was most strongly supported in fall 2019, when dieback increased significantly with decreased elevation and increased southwestness. Aspect was correlated with dieback severity across all years, suggesting this could be a significant variable to consider when generating models for future drought risk. Moreover, it was found that the increase in dieback severity from 2016-2019 was greatest at lower elevations compared to high.
In other words, while most sites experienced an overall increase in dieback, the increase at lower elevations was much more dramatic. Collectively, these results suggest that while populations of A. glauca across the landscape in Santa Barbara county are susceptible to extreme drought, those occurring on exposed, southwest-facing slopes and/or at the lower edge of their range may be less capable of recovery, particularly during prolonged drought. This is consistent with studies showing that low elevations and exposed slopes can correlate with plant water stress and mortality . Interestingly, the dieback severity was greatest in fall 2019, despite the region experiencing above-average rainfall in both the 2016-17 and 2018-19 water years . Thus, a single year of above-average rainfall was not sufficient to restore shrubs to predrought canopy health levels. One possible explanation for this is that 2011-2016 and the 2017-18 wet seasons were extremely dry, pushing A. glauca individuals towards a threshold of drought resistance that could not be counteracted by one or two wet years. Venturas et al. found A. glauca to have high levels of hydraulic failure, dieback, and whole shrub mortality in Malibu County during the drought year of 2014, and a study by Paddock et al. from an intense drought year in California yielded similarly high mortality for this species. This supports the significance of intense drought years that preceded higher rainfall years in our study. Additionally, Gill and Mahall and Mahall et al. found evidence that some chaparral shrubs do not respond to surplus water, and thus high-rainfall years may not be accurate predictors of shrub recovery. Shallow-rooted species like A. glauca may only be able to benefit from soil water availability near the surface, and/or they may be relying on water availability in rocky outcrops that is more dependent on outcrop cavity structure than on individual rain event totals. Therefore, short-term excessive rainfall may be irrelevant for the recovery of these shrubs. Another hypothesis is that opportunistic fungi, as identified in prior work , may have played a significant role in the sudden development and continuation of dieback. Biotic agents are known to exacerbate drought effects by amplifying hydraulic failure and/or carbon starvation . A. glauca, already weakened by drought, procona valencia may have been further impacted by fungal pathogens emerging from their latency between 2012 and 2016. Drake-Schultheis et al. have documented synergistic interactions of drought and latent fungal pathogens on A. glauca in a greenhouse setting. However, more research is required to identify the exact role of latent fungal pathogens in A. glauca in the field. Interestingly, Mahall et al. report detailed phenological measurements of A. glauca individuals in a nearby population in the 1980s following individual leaves for up to three years. They do not report dieback or evidence of fungal disease in the canopy of any individuals despite detailed descriptions of leaf condition. Persona communication with Mahall and Thwing likewise confirm the absence of evidence for pathogenic fungi in their study plants. Drake-Schultheis et al. have evidence for the very recent introduction of the most virulent pathogen in this system, Neofusicoccum australe. While the Mahall et al. study was conducted during the drought of the late 1980s and this drought was similar in magnitude, it may be cumulative drought and a new fungal pathogen together that are today causing canopy dieback.A surprising result of this study was that while mortality among A. glauca was noted across the landscape at the beginning of this study, no new mortality was observed over the following four years of drought, despite some very low XPP measurements in 2016.
Indeed, the XPP obtained in Fall 2016 are almost twice as negative as those measured in A. glauca shrubs in this same region during an earlier drought in the late 1980s . Furthermore, a significant number of shrubs across sites reached extreme levels of canopy dieback yet survived and in some cases even showed evidence of recovery. These observations are indicative of an impressive resilience to drought stress in this chaparral shrub species. As previously reported, A. glauca are typically identified as anisohydric, exhibiting continued gas exchange during drought and high resistance to cavitation . Our data showing the continuation of gas exchange well into the dry season further supports this strategy, although Anet did shut down in the lower elevation populations. The Venturas et al. and Paddock et al. studies also found greater mortality in A. glauca than was recorded in our study, and in other chaparral shrubs with high cavitation resistance– a relationship suggested to be the result of greater susceptibility to high intensity drought. Therefore, why high mortality was not also observed in our study is unclear. It may be a result of Santa Barbara experiencing a slightly milder climate, compared to those of the aforementioned studies, which both took place in more arid regions of southern California, including in a chaparral-desert ecotone. Summer fog, a normal occurrence in Santa Barbara County, may also play an important role in reducing vapor pressure deficit, thus providing critical drought stress relief during the summer months. Furthermore, our study areas were dominated by stands of relatively large, mature plants, which have been shown to exhibit greater survival rates during drought than smaller individuals . Future studies which directly compare climatic variables and mortality of A. glauca – of various ages – across ecoregions during extreme drought would therefore be of great value, and may elucidate a better understanding of the driving factors and relative vulnerabilities of plants in high-risk ecosystems. Consistent with our hypothesis, plants located at lower elevations generally experienced more physiological stress and greater dieback severity than those at higher elevations. However, there was also considerable variability within sites and along the elevational gradient. We hypothesize this is due to the extreme heterogeneity of the Santa Ynez mountain landscape, and that certain landscape features outside the scope of this study may act as refugia for A. glauca resilience. Since A. glauca is a relatively shallowly-rooted species, variables such as slope, aspect, slope angle, concavity, and rockiness – all of which vary greatly across the region and influence temperature, water availability, or both – likely impact shrub functioning and vulnerability to drought on a very local scale. In considering long-term predictions, it is possible that while some stands of A. glauca may suffer high levels of dieback and even mortality during extreme drought, populations as a whole may be resilient because of microsite and landscape heterogeneity.Although higher elevations in our study area historically record greater rainfall, we found this trend to be slightly disrupted during our study period: rainfall at the highest elevation site was lower than at the intermediate site from 2015-16 to 2018-19 . Furthermore, while elevation is known to influence temperature such that lower elevations are generally hotter than higher ones within a region, human-induced climate warming has been shown in some cases to occur more rapidly at higher than at lower elevations. For instance, Giambelluca et al. found more rapid warming at elevations above 800m in Hawaii, and a review by Pepin et al. found enhanced warming with elevation in high mountain regions globally . Therefore, higher elevation sites in our study may also be experiencing more warming. We do not have temperature records for our six sites so we cannot separate temperature trends or anomalies from other elevational impacts. Nonetheless, the lack of a consistent elevation effect could be due not only to the confounding effects of other topographic features , but also to lower rainfall and increased warming in some higher elevation sites. From the early studies of Whittaker ecologists have long recognized the influence of aspect on plant composition and vegetation dynamics in montane environments.