Frosty pod rot disease of cacao (FPR), caused by the fungus Moniliophthora roreri, has severely impacted the production of cocoa in Latin America since its discovery. Prior to the 1950s, FPR was known only from Colombia and Ecuador. However, beginning in the 1970s, its geographical range has dramatically expanded throughout most of the cacao‐producing regions of the Americas. The origin of the pathogen remains unknown. In this study, we evaluated the genetic diversity of M. roreri from areas spanning, as much as possible, its current geographical range using simple‐sequence repeat markers and a publicly available single‐nucleotide polymorphism data set. Two hotspots of genetic diversity were found: coastal Ecuador and the inter‐Andean Magdalena Valley of Colombia, neither of which correspond to the Amazonian origin of the host. However, both areas were early centres of intense cultivation of cacao. Our results indicate that M. roreri was introduced into both areas from its centre of origin, where intensive cacao cultivation probably led to the increase of inoculum and further dissemination of the disease. Current invasions can be traced to two genotypes responsible for all known instances of the pathogen in Central America, the Caribbean, Peru and Bolivia. We also report for the first time M. roreri in Maynas (Peruvian Amazon), which is probably the result of a recent introduction from Colombia.

A common issue in forest management is related to yield projection for stands at young ages. This study aimed to evaluate the Clutter model and artificial neural networks for projecting eucalypt stands production from early ages, using different data arrangements. In order to do this, the changes in the number of measurement intervals used as input in the Clutter model and artificial neural networks (ANNs) are tested. The Clutter model was fitted considering two sets of data: usual, with inventory measurements (I) paired at intervals each year (I1–I2, I2–I3, …, In–In+1); and modified, with measurements paired at all possible age intervals (I1–I2, I1–I3, …, I2–I3, I2–I4, …, In–In+1). The ANN was trained with the modified dataset plus soil type and geographic coordinates as input variables. The yield projections were made up to the final ages of 6 and 7 years from all possible initial ages (2, 3, 4, 5, or 6 years). The methods are evaluated using the relative error (RE%), bias, correlation coefficient (ryŷ), and relative root mean square error (RMSE%). The ANN was accurate in all cases, with RMSE% from 8.07 to 14.29%, while the Clutter model with the modified dataset had values from 7.95 to 23.61%. Furthermore, with ANN, the errors were evenly distributed over the initial projection ages. This study found that ANN had the best performance for stand volume projection surpassing the Clutter model regardless of the initial or final age of projection.

No abstract

No abstract

<span class="fontstyle0">La estimación precisa del volumen del vuelo forestal hace posible estimar un valor monetario correcto. El objetivo de este estudio fue valorar el vuelo forestal del genero </span><span class="fontstyle2">Pinus </span><span class="fontstyle0">en el Parque Forestal Aylambo de la Universidad Nacional de Cajamarca, aplicando el modelo volumétrico de Schumacher y Hall (1993); con la finalidad de comparar la valoración económica con la valoración económica ajustada de la plantación, y conocer el valor actual del vuelo forestal. Se realizó un inventario al 100%. El modelo volumétrico fue ajustado utilizando regresión lineal, a través del método de mínimos cuadrados. Presentando la especie </span><em><span class="fontstyle2">Pinus patula </span></em><span class="fontstyle0">un R</span><span class="fontstyle0">2 </span><span class="fontstyle0">de 78,95% y R</span><span class="fontstyle0">2ajustado </span><span class="fontstyle0">de 78,83% y para la especie de </span><em><span class="fontstyle2">Pinus radiata </span></em><span class="fontstyle0">un R</span><span class="fontstyle0">2 </span><span class="fontstyle0">de 85,22% y R</span><span class="fontstyle0">2 ajustado </span><span class="fontstyle0">de 84,9%. Para estimar el volumen ajustado en m</span><span class="fontstyle0">3</span><span class="fontstyle0">, en razón al diámetro a la altura del pecho (DAP) y la altura comercial (Hc) fueron tomados en metros; los modelos determinados son: LnV = -0,9347984 + 2,00002362*LnDAP + 1,0015779*LnHc para </span><em><span class="fontstyle2">P. patula </span></em><span class="fontstyle0">y LnV = 0,35264369 + 1,9401478*LnDAP + 1,2038246*LnHc para </span><em><span class="fontstyle2">P. radiata</span></em><span class="fontstyle0">. El valor del vuelo forestal de la plantación fue de 4351,11 nuevos soles y el valor ajustado de 4344,71 nuevos soles.</span>

<p>El presente estudio fue realizado con los objetivos de regular la producción forestal con pronósticos de modelación Clutter y Redes Neuronales Artificiales (RNA) en plantaciones de eucalipto (híbridos de Eucalyptus urophylla x Eucalyptus grandis), localizado en la región Centro Oeste del Estado de Minas Gerais, Brasil. De las tablas de producción se generaron procesos de regulación forestal en formato de programación lineal utilizando el software RPF 2.0®, con el fin de maximizar el ingreso líquido futuro para 22 años, 2016 – 2038. Se generaron dos planos de manejo, obteniendo variables de prescripción, edad, rotación, volumen, costos, demanda. Para la aplicación del modelo de programación lineal se consideró un horizonte de planeamiento de 22 periodos, tasa de interés de 11,5% y una edad regulatoria de 7 años con 0% de variación permitida, considerando una sola rotación y una variación de producción entre 100 000 m3 y 60 000 m3. Se concluye que los dos planos de proyección tienen influencia en la regulación forestal, destacando la técnica de RNA como una alternativa más real y confiable para regular una plantación.</p>

The site quality of plantations is estimated mainly by the local index used in forest management for growth and production models. The indices are obtained by mathematical models using variable Height dominant (Hd), however to obtain the measurements of this variable, its demands many economic resources at permanent plots inventories. In the present research, it was determined the efficiency of five Analytical Methods usage in the local Index estimation using the dominant Height (Hd) and the dominant Dap variable (Dapd) by a comparison in the local index station of the plots; being adjusted the models with the Software Statistica. The local indices method is the one that has given the best result, preliminarily defined with 23 % variation in the classification of plots as a function of the classification with the variable Hd. This method was adjusted with the Richards model, obtaining a coefficient of correlation between the observed data and the estimated 0,982. The dominant Dap variable can be used in the estimation of site quality.

RESUMENEl presente estudio fue realizado con los objetivos de comparar, ajustar y evaluar el modelo de crecimiento y producción propuesto por Clutter (Sistema de Ecuaciones) y por Redes Neuronales Artificiales (RNA) para una plantación de eucalipto (híbridos de Eucalyptus urophylla x Eucalyptus grandis), localizado en la región Centro Oeste del Estado de Minas Gerais, Brasil. Medidas en diferentes años, entre los 19 y 106 meses. Se realizó cinco mediciones sucesivas. Después del análisis, se verificó que el algoritmo computacional RNA, es más preciso en las estimaciones que la regresión lineal con la ventaja de agregar datos categóricos y el menor tiempo de procesamiento de datos.Palabras clave: Eucalipto, redes neuronales artificiales, RNA, sistema de ecuaciones. ABSTRACTThe present study was carried out with the objective of comparing, adjusting and evaluating the growth and production model proposed by Clutter (System of Equations) and Artificial Neural Networks (ANN) for a plantation of eucalyptus (Eucalyptus urophylla x Eucalyptus grand is hybrids ). Located in the Midwest region of the State of Minas Gerais, Brazil, measured in different years between 19 and 106 months, with at least five successive measurements. After the analysis, it was verified that using the computational algorithm ANN proved to be more accurate in the estimations than using linearregression, having the advantage of adding categorical data and shorter data processing time.