Search

In arid and semiarid systems of western North America, the most damaging invasive plants are winter annuals. These plants are destroying wildlife habitat, reducing livestock production, and increasing wildfires. Monitoring these plants for lasting population changes is challenging because their abundances vary widely from year to year. Some of this variation is due to weather, and quantifying effects of weather is important for distinguishing transcient from lasting population changes and understanding effects of climate change. Fall and spring weather affect germination and seed production of the current generation of plants and, therefore, impact population sizes of subsequent generations of plants. Extensive data are required to estimate effects of fall and spring weather on multiple generations of plants. We used Bayesian statistics to integrate experimental and long‐term (31 years) monitoring data and quantify invasive annual grass [downy brome (Bromus tectorum L.) and Japanese brome (Bromus japonicus Thunb.)] responses to weather. Bromes ranged from nearly absent to comprising half of total biomass depending on three previous years of weather. Brome biomass increased with precipitation one, two, and three falls prior to measurement. Fall precipitation is projected to increase, and a mere 6.5 mm increase, which is just 2% of mean annual precipitation, would increase brome biomass 40% (28%, 54%) (mean [95% CI]) according to our model. Increased fall precipitation could favor many invasive winter annual grasses and forbs. Dry spring conditions reduced brome biomass the current year but increased brome biomass one and likely two (p = 0.08) years later, perhaps because dry conditions weakened perennial competitors. This finding casts doubt on several one‐year precipitation experiments that concluded drier spring weather would reduce brome abundances. Integrating short‐term experiments and long‐term monitoring is useful for estimating invasive plant responses to the weather and characterizing their responses to climate change. Our research provides predictions of brome abundances that could improve monitoring efforts by helping land managers interpret population dynamics in the context of seasonal precipitation patterns. [ABSTRACT FROM AUTHOR]

Accurate aboveground vegetation biomass forecasting is essential for livestock management, climate impact assessments, and ecosystem health. While artificial intelligence (AI) techniques have advanced time series forecasting, a research gap in predicting aboveground biomass time series beyond single values persists. This study introduces RECMO and DirRecMO, two multi-output methods for forecasting aboveground vegetation biomass. Using convolutional neural networks, their efficacy is evaluated across short-, medium-, and long-term horizons on six Kenyan grassland biomass datasets, and compared with that of existing single-output methods (Recursive, Direct, and DirRec) and multi-output methods (MIMO and DIRMO). The results indicate that single-output methods are superior for short-term predictions, while both single-output and multi-output methods exhibit a comparable effectiveness in long-term forecasts. RECMO and DirRecMO outperform established multi-output methods, demonstrating a promising potential for biomass forecasting. This study underscores the significant impact of multi-output size on forecast accuracy, highlighting the need for optimal size adjustments and showcasing the proposed methods' flexibility in long-term forecasts. Short-term predictions show less significant differences among methods, complicating the identification of the best performer. However, clear distinctions emerge in medium- and long-term forecasts, underscoring the greater importance of method choice for long-term predictions. Moreover, as the forecast horizon extends, errors escalate across all methods, reflecting the challenges of predicting distant future periods. This study suggests advancing hybrid models (e.g., RECMO and DirRecMO) to improve extended horizon forecasting. Future research should enhance adaptability, investigate multi-output impacts, and conduct comparative studies across diverse domains, datasets, and AI algorithms for robust insights. [ABSTRACT FROM AUTHOR]

Timely forecasting of aboveground vegetation biomass is crucial for effective management and ensuring food security. However, research on predicting aboveground biomass remains scarce. Artificial intelligence (AI) methods could bridge this research gap and provide early warning to planners and stakeholders. This study evaluates the effectiveness of deep learning (DL) algorithms in predicting aboveground vegetation biomass with limited-size data. It employs an iterative forecasting procedure for four target horizons, comparing the performance of DL models—multi-layer perceptron (MLP), long short-term memory (LSTM), gated recurrent unit (GRU), convolutional neural network (CNN), and CNN-LSTM—against the traditional seasonal autoregressive integrated moving average (SARIMA) model, serving as a benchmark. Five limited-size vegetation biomass time series from Kenyan grasslands with values at 15-day intervals over a 20-year period were chosen for this purpose. Comparing the outcomes of these models revealed significant differences (p < 0.05); however, none of the models proved superior among the five time series and the four horizons evaluated. The SARIMA, CNN, and CNN-LSTM models performed best, with the statistical model slightly outperforming the other two. Additionally, the accuracy of all five models varied significantly according to the prediction horizon (p < 0.05). As expected, the accuracy of the models decreased as the prediction horizon increased, although this relationship was not strictly monotonic. Finally, this study indicated that, in limited-size aboveground vegetation biomass time series, there is no guarantee that deep learning methods will outperform traditional statistical methods. [ABSTRACT FROM AUTHOR]

Livestock production in semi-arid areas has been unpredictable due to climate variability, mainly rainfall. This study aims to simulate rangeland production variability aff ected by rainfall over time and relate it to an adjusted carrying capacity, using forage stock to maximize the potential of production of the system to a specific guarantee level. Regression analysis of forage biomass against rainfall was performed for ecological sites in the Brazilian semi-arid region to generate probability distribution curves for the historical rainfall for each location using Monte Carlo approach. Forage biomass variability estimated over time was used as input to the model. The system optimizes forage use at a sustainable stocking rate and uses forage surpluses in good years to fill deficits during adverse years, due to a certain level of guarantee. As a rule, smallholder farmers would need to maintain a storage of around 1,500 kg ha-1 of DM of forage to maintain an adjusted carrying capacity of 0.11 animal units ha-1, with a guarantee of 95% in the long term, stressing the forage storage capacity as a central component of the model. Since farm size influences forage production capacity and mainly forage stock capacity, recommendations to cope with this paradigm are suggested. [ABSTRACT FROM AUTHOR]

Globally, rangelands face interacting pressures from climate, land-use, socio-economic and political changes, all of which threaten herder livelihoods and grassland health. Given these dynamics, it is often unclear which policies would best support sustainable land use and livelihoods in the future. There are multiple theories for how tenure, rules, social relations and environmental variability intersect to influence pastoral resource governance, but these have largely been developed based on empirical data from specific social and environmental contexts. Few studies have attempted to evaluate empirically how the factors driving pastoral management decisions might vary across a social-ecological gradient. In this work we attempt to reconcile the current diversity of theories around pastoral resource governance with an empirical dataset on household herd and pasture management decisions in Mongolia, where there is ongoing debate over proposed rangeland policies, including formalization of land tenure. We assess the relationship between theorized predictors of herder behavior (i.e. formal rights, formal rules, social capital, and environmental variability), and household-level management decisions about pastoral mobility and storage. We compare our findings from a survey of 760 households across four ecozones to predictions from pastoral resource governance theories to assess which theories best match the complex realities on the ground. We observed a continuum of de facto pastoral governance regimes that roughly map onto a social-ecological gradient defined by (i) resource variability and predictability, and (ii) the relative prevalence of formal rule-based vs. implicit norm-based governance. We find that rules, social capital, and forage availability are the strongest predictors of whether a household will reserve forage, with consistent effects across ecological zones. In contrast, social ties and environmental conditions most strongly predict mobility practices. While most households reserved pastures regardless of tenure status, those households who do not reserve pastures are more likely to lack formal use rights. Our findings reinforce the importance of avoiding "one size fits all" rangeland governance policies. We demonstrate that herders make decisions in response to environmental productivity and variability over space and time, and that social ties and mobility are critical for maintaining access to forage. This may explain why many herders remain skeptical of formal pasture tenure, which could restrict this flexibility. Policies that ensure herders' collective rights to pasture through large-scale zoning rather than tenure could potentially achieve both secure pasture rights and maximum flexibility. • We compare competing theories about the governance of resources in pastoral systems. • We test these theories with an empirical dataset from Mongolia, where there is ongoing debate over proposed rangeland policies, including formalization of land tenure. • Rules, social capital, and forage availability influence resource storage more than formalized tenure. • Social relations are key for maintaining mobility practices. • We describe a continuum of de facto pastoral governance regimes along a social-ecological gradient. • Findings reinforce the importance of avoiding "one size fits all" rangeland governance policies. [ABSTRACT FROM AUTHOR]

Despite their size, small farm ponds are important features in many landscapes. Yet hydrographical databases often fail to capture these ponds, and their impacts on watershed processes remain unclear. For a 230-km2 portion of central Texas, United States ( U.S.), we created a historical inventory of ponds and quantified the accuracy of automated detection methods under varying drought conditions. In addition, we documented pond dredging/enlargement events and identified sites that had been abandoned. We also analyzed sediment cores from downstream reservoirs to track changes in watershed sediment transport. Over 75 years, pond densities increased more than 350% - to among the highest documented in the U.S. - and the ability of automated methods to detect these ponds was highly dependent on drought severity ( R2 = 0.96). Approximately 5% of ponds present in the 1950s were no longer present in 2012, while 33% were dredged between 1937 and 2012. Downstream reservoir sedimentation has decreased by an average of 55% as ponds have increased in number. These findings suggest that small ponds and the maintenance of trapping efficiency have large-scale impacts on sediment dynamics. Accurately accounting for these storage effects is vital to water resource planning efforts. [ABSTRACT FROM AUTHOR]

Woody plant encroachment-the conversion of grasslands to woodlands-continues to transform rangelands worldwide, yet its causes and consequences remain poorly understood. Despite this being a coupled human-ecological phenomenon, research to date has tended toward ecological aspects of the issue. In this paper, we provide new insight into the long-term relationships between human demographics and woody plant cover at the landscape scale. We used time-series aerial imagery and historical census data to quantify changes in population, land ownership patterns, and woody cover between 1937 and 2012 in three different settings in central Texas, USA. Woody cover closely paralleled population in a semi-urban watershed ( R2 = 0.81) and two separate clusters of rural watersheds ( R2 = 0.88 and 0.93), despite exhibiting very different directional trends over time in each setting. Woody cover also closely tracked average farm size in each rural watershed cluster ( R2 = 0.57 and 0.90). These results highlight a tight coupling between demographic trends and the extent of woody plant cover. Such human factors may explain a great deal of woody plant cover patterns in other global rangeland systems with similar historical contexts and serve as a predictive proxy of landscape trends. Accordingly, policy recommendations should consider these demographic factors, and future woody plant encroachment research should explicitly include human dimensions. [ABSTRACT FROM AUTHOR]