Statistical modelling of determinants of child stunting using secondary data and Bayesian networks: a UKRI Global Challenges Research Fund (GCRF) Action Against Stunting Hub protocol paper.

INTRODUCTION: Several factors have been implicated in child stunting, but the precise determinants, mechanisms of action and causal pathways remain poorly understood. The objective of this study is to explore causal relationships between the various determinants of child stunting. METHODS AND ANALYSIS: The study will use data compiled from national health surveys in India, Indonesia and Senegal, and reviews of published evidence on determinants of child stunting. The data will be analysed using a causal Bayesian network (BN)-an approach suitable for modelling interdependent networks of causal relationships. The model's structure will be defined in a directed acyclic graph and illustrate causal relationship between the variables (determinants) and outcome (child stunting). Conditional probability distributions will be generated to show the strength of direct causality between variables and outcome. BN will provide evidence of the causal role of the various determinants of child stunning, identify evidence gaps and support in-depth interrogation of the evidence base. Furthermore, the method will support integration of expert opinion/assumptions, allowing for inclusion of the many factors implicated in child stunting. The development of the BN model and its outputs will represent an ideal opportunity for transdisciplinary research on the determinants of stunting. ETHICS AND DISSEMINATION: Not applicable/no human participants included.

Poverty dynamics and the determining factors among East African smallholder farmers.

CONTEXT: Rapid economic development in East Africa is matched by extremely dynamic smallholder livelihoods. Objective: To quantify the changes in poverty of smallholder farmers, to evaluate the potential of farm and off-farm activities to alleviate poverty, and to evaluate the potential barriers to poverty alleviation. METHODS: The analyses were based on a panel survey of 600 households undertaken in 2012 and re-visited approximately four years later in four sites in East Africa. The sites represented contrasting smallholder farming systems, linked to urban centres undergoing rapid economic and social change (Nairobi, Kampala, Kisumu, and Dar-es-Salaam). The surveys assessed farm management, farm productivity, livelihoods, and various measures of household welfare. RESULTS AND CONCLUSIONS: Almost two thirds of households rose above or fell below meaningful poverty thresholds - more than previously measured in this context - but overall poverty rates remained constant. Enhanced farm value production and off-farm income proved to be important mechanisms to rise out of poverty for households that were already resource-endowed. However, households in the poorest stratum in both panels appeared to be stuck in a poverty trap. They owned significantly fewer productive assets in the first panel compared to other groups (land and livestock), and these baseline assets were found to be positively correlated with farm income in the second panel survey. Equally these households were also found to be among the least educated, while education was found to be an important enabling factor for the generation of high value off-farm income. SIGNIFICANCE: Rural development that aims to stimulate increases in farm produce value as a means to alleviate poverty are only viable for already resource-endowed households, as they have the capacity to enhance farm value production. Conversely, the alleviation of extreme poverty should focus on different means, perhaps cash transfers, or the development of more sophisticated social safety nets. Furthermore, while off-farm income presents another important mechanism for poverty alleviation in rural areas, these opportunities are restricted to those households that have had access to education. As more households turn to off-farm activities to supplement or replace their livelihoods, farming approaches will also change affecting the management of natural resources. These dynamics ought to be better understood to better manage land-use transitions.

Nutritional Quality, Voluntary Intake and Enteric Methane Emissions of Diets Based on Novel Cayman Grass and Its Associations With Two Leucaena Shrub Legumes.

Methane (CH4) emissions from enteric fermentation in cattle are an important source of greenhouse gases, accounting for about 40% of all agricultural emissions. Diet quality plays a fundamental role in determining the magnitude of CH4 emissions. Specifically, the inclusion of feeds with high digestibility and nutritional value have been reported to be a viable option for reducing CH4 emissions and, simultaneously, increase animal productivity. The present study aimed to evaluate the effect of the nutritional composition and voluntary intake of diets based on tropical forages upon CH4 emissions from zebu steers. Five treatments (diets) were evaluated: Cay1: Urochloa hybrid cv. Cayman (harvested after 65 days of regrowth: low quality); Cay2: cv. Cayman harvested after 45 days of regrowth; CayLl: cv. Cayman + Leucaena leucocephala; CayLd: cv. Cayman + Leucaena diversifolia; Hay: Dichantium aristatum hay as a comparator of common naturalized pasture. For each diet representing different levels of intensification (naturalized pasture, improved pasture, and silvopastoral systems), CH4 emissions were measured using the polytunnel technique with four zebu steers housed in individual chambers. The CH4 accumulated was monitored using an infrared multigas analyzer, and the voluntary forage intake of each animal was calculated. Dry matter intake (DMI, % of body weight) ranged between 0.77 and 2.94 among diets offered. Emissions of CH4 per kg of DMI were significantly higher (P < 0.0001) in Cay1 (60.4 g), compared to other treatments. Diets that included Leucaena forage legumes had generally higher crude protein contents and higher DMI. Cay1 and Hay which had low protein content and digestibility had a higher CH4 emission intensity (per unit live weight gain) compared to Cay2, CayLl and CayLd. Our results suggest that grass consumed after a regrowth period of 45 days results in lower CH4 emissions intensities compared to those observed following a regrowth period of 65 days. Diets with Leucaena inclusion showed advantages in nutrient intake that are reflected in greater live weight gains of cattle. Consequently, the intensity of the emissions generated in the legume-based systems were lower suggesting that they are a good option for achieving the emission reduction goals of sustainable tropical cattle production.

Evidence-based investment selection: Prioritizing agricultural development investments under climatic and socio-political risk using Bayesian networks.

Agricultural development projects have a poor track record of success mainly due to risks and uncertainty involved in implementation. Cost-benefit analysis can help allocate resources more effectively, but scarcity of data and high uncertainty makes it difficult to use standard approaches. Bayesian Networks (BN) offer a suitable modelling technology for this domain as they can combine expert knowledge and data. This paper proposes a systematic methodology for creating a general BN model for evaluating agricultural development projects. Our approach adapts the BN model to specific projects by using systematic review of published evidence and relevant data repositories under the guidance of domain experts. We evaluate a large-scale agricultural investment in Africa to provide a proof of concept for this approach. The BN model provides decision support for project evaluation by predicting the value-measured as net present value and return on investment-of the project under different risk scenarios.

The Rural Household Multiple Indicator Survey, data from 13,310 farm households in 21 countries.

The Rural Household Multiple Indicator Survey (RHoMIS) is a standardized farm household survey approach which collects information on 758 variables covering household demographics, farm area, crops grown and their production, livestock holdings and their production, agricultural product use and variables underlying standard socio-economic and food security indicators such as the Probability of Poverty Index, the Household Food Insecurity Access Scale, and household dietary diversity. These variables are used to quantify more than 40 different indicators on farm and household characteristics, welfare, productivity, and economic performance. Between 2015 and the beginning of 2018, the survey instrument was applied in 21 countries in Central America, sub-Saharan Africa and Asia. The data presented here include the raw survey response data, the indicator calculation code, and the resulting indicator values. These data can be used to quantify on- and off-farm pathways to food security, diverse diets, and changes in poverty for rural smallholder farm households.

Adequate vegetative cover decreases nitrous oxide emissions from cattle urine deposited in grazed pastures under rainy season conditions.

A decline in pasture productivity is often associated with a reduction in vegetative cover. We hypothesize that nitrogen (N) in urine deposited by grazing cattle on degraded pastures, with low vegetative cover, is highly susceptible to losses. Here, we quantified the magnitude of urine-based nitrous oxide (N2O) lost from soil under paired degraded (low vegetative cover) and non-degraded (adequate vegetative cover) pastures across five countries of the Latin America and the Caribbean (LAC) region and estimated urine-N emission factors. Soil N2O emissions from simulated cattle urine patches were quantified with closed static chambers and gas chromatography. At the regional level, rainy season cumulative N2O emissions (3.31 versus 1.91 kg N2O-N ha-1) and emission factors (0.42 versus 0.18%) were higher for low vegetative cover compared to adequate vegetative cover pastures. Findings indicate that under rainy season conditions, adequate vegetative cover through proper pasture management could help reduce urine-induced N2O emissions from grazed pastures.

Strengths and limitations of computer assisted telephone interviews (CATI) for nutrition data collection in rural Kenya.

Despite progress in fighting undernutrition, Africa has the highest rates of undernutrition globally, exacerbated by drought and conflict. Mobile phones are emerging as a tool for rapid, cost effective data collection at scale in Africa, as mobile phone subscriptions and phone ownership increase at the highest rates globally. To assess the feasibility and biases of collecting nutrition data via computer assisted telephone interviews (CATI) to mobile phones, we measured Minimum Dietary Diversity for Women (MDD-W) and Minimum Acceptable Diet for Infants and Young Children (MAD) using a one-week test-retest study on 1,821 households in Kenya. Accuracy and bias were assessed by comparing individual scores and population prevalence of undernutrition collected via CATI with data collected via traditional face-to-face (F2F) surveys. We were able to reach 75% (n = 1366) of study participants via CATI. Women's reported nutrition scores did not change with mode for MDD-W, but children's nutrition scores were significantly higher when measured via CATI for both the dietary diversity (mean increase of 0.45 food groups, 95% confidence interval 0.34-0.56) and meal frequency (mean increase of 0.75 meals per day, 95% confidence interval 0.53-0.96) components of MAD. This resulted in a 17% higher inferred prevalence of adequate diets for infants and young children via CATI. Women without mobile-phone access were younger and had fewer assets than women with access, but only marginally lower dietary diversity, resulting in a small non-coverage bias of 1-7% due to exclusion of participants without mobile phones. Thus, collecting nutrition data from rural women in Africa with mobile phones may result in 0% (no change) to as much as 25% higher nutrition estimates than collecting that information in face-to-face interviews.

Spatial patterns of tree yield explained by endogenous forces through a correspondence between the Ising model and ecology.

Spatial patterning of periodic dynamics is a dramatic and ubiquitous ecological phenomenon arising in systems ranging from diseases to plants to mammals. The degree to which spatial correlations in cyclic dynamics are the result of endogenous factors related to local dynamics vs. exogenous forcing has been one of the central questions in ecology for nearly a century. With the goal of obtaining a robust explanation for correlations over space and time in dynamics that would apply to many systems, we base our analysis on the Ising model of statistical physics, which provides a fundamental mechanism of spatial patterning. We show, using 5 y of data on over 6,500 trees in a pistachio orchard, that annual nut production, in different years, exhibits both large-scale synchrony and self-similar, power-law decaying correlations consistent with the Ising model near criticality. Our approach demonstrates the possibility that short-range interactions can lead to long-range correlations over space and time of cyclic dynamics even in the presence of large environmental variability. We propose that root grafting could be the common mechanism leading to positive short-range interactions that explains the ubiquity of masting, correlated seed production over space through time, by trees.

Multi-scale measurements show limited soil greenhouse gas emissions in Kenyan smallholder coffee-dairy systems.

Efforts have been made in recent years to improve knowledge about soil greenhouse gas (GHG) fluxes from sub-Saharan Africa. However, data on soil GHG emissions from smallholder coffee-dairy systems have not hitherto been measured experimentally. This study aimed to quantify soil GHG emissions at different spatial and temporal scales in smallholder coffee-dairy farms in Murang'a County, Central Kenya. GHG measurements were carried out for one year, comprising two cropping seasons, using vented static chambers and gas chromatography. Sixty rectangular frames were installed on two farms comprising the three main cropping systems found in the area: 1) coffee (Coffea arabica L.); 2) Napier grass (Pennisetum purpureum); and 3) maize intercropped with beans (Zea mays and Phaseolus vulgaris). Within these fields, chambers were allocated on fertilised and unfertilised locations to capture spatial variability. Cumulative annual fluxes in coffee plots ranged from 1 to 1.9kgN2O-Nha-1, 6.5 to 7.6MgCO2-Cha-1 and -3.4 to -2.2kgCH4-Cha-1, with 66% to 94% of annual GHG fluxes occurring during rainy seasons. Across the farm plots, coffee received most of the N inputs and had 56% to 89% higher emissions of N2O than Napier grass, maize and beans. Within farm plots, two to six times higher emissions were found in fertilised hotspots - around the perimeter of coffee trees or within planted maize rows - than in unfertilised locations between trees, rows and planting holes. Background and induced soil N2O emissions from fertiliser and manure applications in the three cropping systems were lower than hypothesized from previous studies and empirical models. This study supplements methods and underlying data for the quantification of GHG emissions at multiple spatial and temporal scales in tropical, smallholder farming systems. Advances towards overcoming the dearth of data will facilitate the understanding of synergies and tradeoffs of climate-smart approaches for low emissions development.

Nitrous Oxide and Methane Fluxes from Urine and Dung Deposited on Kenyan Pastures.

Livestock keeping is ubiquitous in tropical Africa. Urine and dung from livestock release greenhouse gases (GHGs), such as nitrous oxide (NO) and methane (CH), to the atmosphere. However, the extent of GHG's impact is uncertain due to the lack of in situ measurements in the region. Here we measured NO and CH emissions from cow urine and dung depositions in two Kenyan pastures that received different amounts of rainfall using static chambers across wet and dry seasons. Cumulative NO emissions were greater under dung+urine and urine-only patches ( 0.0001), more than three times higher in the wet compared with the dry season ( 0.0001), and higher in the farm receiving higher rainfall overall ( 0.0001). Cumulative CH emissions differed across treatments ( = 0.012), driven primarily by soil CH uptake from the urine-only treatment. Cumulative NO emissions were positively related to N input rate in excreta. However, the relationship was linear during the dry season ( 0.99; 0.001) and exponential during the wet season ( 0.99; < 0.0001). Nitrous oxide emission factors were 0.05% (dry season) and 0.18% (wet season) of N in urine and dung+urine, which is less than 10% of the IPCC Default Tier 1 emission factor of 2%. We predict that emissions from cattle urine in Kenya are approximately 1.7 Gg NO-N yr (FAO estimates 11.9 Gg NO-N yr). Our findings suggest that current estimates may overestimate the contribution of excreta to national GHG emissions and that emission factors from urine and dung need to account for agroecosystems with distinct wet and dry seasons.

Limits of agricultural greenhouse gas calculators to predict soil N2O and CH4 fluxes in tropical agriculture.

Demand for tools to rapidly assess greenhouse gas impacts from policy and technological change in the agricultural sector has catalyzed the development of 'GHG calculators'- simple accounting approaches that use a mix of emission factors and empirical models to calculate GHG emissions with minimal input data. GHG calculators, however, rely on models calibrated from measurements conducted overwhelmingly under temperate, developed country conditions. Here we show that GHG calculators may poorly estimate emissions in tropical developing countries by comparing calculator predictions against measurements from Africa, Asia, and Latin America. Estimates based on GHG calculators were greater than measurements in 70% of the cases, exceeding twice the measured flux nearly half the time. For 41% of the comparisons, calculators incorrectly predicted whether emissions would increase or decrease with a change in management. These results raise concerns about applying GHG calculators to tropical farming systems and emphasize the need to broaden the scope of the underlying data.

Plant reproduction and environmental noise: how do plants do it?

Plant populations exhibit a wide continuum of reproductive behavior, ranging from nearly constant reproductive output on one end to the extreme of masting (synchronized, highly variable reproduction) on the other. Here, we show that including variability (noise) in density-dependent pollen limitation in current models for pollen-limited plant reproduction may produce any behavior on this continuum. We previously showed that (large) variability in pollination efficiency (a related phenomenon) may induce masting in non-pollen-limited plant populations. Other modeling studies have shown that including variability in accumulated resources (and/or the threshold for reproduction) may induce masting, but do account for masting in non-pollen-limited plant populations. Thus, our results suggest that the range of plant reproductive behavior may be explained with the simple resource budget model combined with the biological realism of variability in density-dependent pollen limitation. This is a specific example of an important functional consequence of the interactions between stochasticity and nonlinearity, and highlights the importance of carefully considering both the biological basis and the mathematical effects of the noise term.

Agriculture's Contribution to Nitrate Contamination of Californian Groundwater (1945-2005).

Nitrogen (N) use in intensive agriculture can degrade groundwater resources. However, considerable time lags between groundwater recharge and extraction complicate source attribution and remedial responses. We construct a historic N mass balance of two agricultural regions of California to understand trends and drivers of past and present N loading to groundwater (1945-2005). Changes in groundwater N loading result from historic changes in three factors: the extent of agriculture (cropland area and livestock herd increased 120 and 800%, respectively), the intensity of agriculture (synthetic and manure waste effluent N input rates increased by 525 and 1500%, respectively), and the efficiency of agriculture (crop and milk production per unit of N input increased by 25 and 19%, respectively). The net consequence has been a greater-than-order-of-magnitude increase in nitrate (NO) loading over the time period, with 163 Gg N yr now being leached to groundwater from approximately 1.3 million ha of farmland (not including alfalfa [ L.]). Meeting safe drinking water standards would require NO leaching reductions of over 70% from current levels through reductions in excess manure applications, which accounts for nearly half of all groundwater N loading, and through synthetic N management improvements. This represents a broad challenge given current economic and technical conditions of California farming if farm productivity is to be maintained. The findings illustrate the growing tension-characteristic of agricultural regions globally-between intensifying food, feed, fiber, and biofuel production and preserving clean water.

The role of large environmental noise in masting: general model and example from pistachio trees.

Masting is synchronous, highly variable reproduction in a plant population, or synchronized boom-bust cycles of reproduction. These pulses of resources have cascading effects through ecosystems, and thus it is important to understand where they come from. How does masting happen and synchronize? In this paper, we suggest a mechanism for this. The mechanism is inspired by data from a pistachio orchard, which suggest that large environmental noise may play a crucial role in inducing masting in plant populations such as pistachio. We test this idea through development and analysis of a mathematical model of plant reproduction. We start with a very simple model, and generalize it based on the current models of plant reproduction and masting. Our results suggest that large environmental noise may indeed be a crucial part of the mechanism of masting in certain types of plant populations, including pistachio. This is a specific example of an important functional consequence of the interactions between stochasticity and nonlinearity.