|AUCLAIR, ALLAN - Animal And Plant Health Inspection Service (APHIS)|
|Perez De Leon, Adalberto - Beto|
|TEEL, PETE - Texas A&M University|
|MESSENGER, MATTHEW - Animal And Plant Health Inspection Service (APHIS)|
|BONILLA, DENISE - Animal And Plant Health Inspection Service (APHIS)|
Submitted to: Ag Data Commons
Publication Type: Database / Dataset
Publication Acceptance Date: 5/6/2020
Publication Date: 5/6/2020
Citation: Auclair, A.N., Perez De Leon, A.A., Teel, P.D., Manoukis, N., Messenger, M.T., Bonilla, D.L. 2020. Novel hurricane hypothesis predicts U.S Cattle Fever Tick outbreaks. Data. https://doi.org/10.15482/USDA.ADC/1518654.
Interpretive Summary: This dataset includes data from 1959-2017 on Cattle Fever Tick (CFT) abundances in the lower Rio Grande Valley of Texas. The package includes an analysis relating abundances of this cattle disease vector against climatic variables. The dataset was used to test a hypothesis relating hurricane activity with CFT abundance.
Technical Abstract: Data Sources: Time series data on cattle fever tick incidence, 1959-2017, and climate variables January 1950 through December 2017, form the core information in this analysis. All variables are monthly averages or sums over the fiscal year (FY), October 01 (of the prior calendar year, y-1) through September 30 of the current calendar year (y). Annual records on detections of Rhipicephalus (Boophilus) microplus and R. (B). annulatus (cattle fever tick, CFT) on premises within the Permanent Quarantine Zone (PQZ) were obtained from CFTEP (USDA-APHIS and the USDA- ARS). Details of tick survey procedures, CFTEP program goals and history, and the geographic extent of the Permanent Quarantine Zone are in SI (Introduction, details). Solar radio flux data as well as Pacific Ocean El Niño Oscillation index data, 1950-2017, are accessed at from NOAA ESRL (2018b). Predicted values for on going Solar Cycle 24 are from NOAA SWPC (2018). Accumulated Cyclone Energy Index (ACE) data are from the NOAA ESRL (2018a) database. Hurricane incidence data over the PQZ are accessed at the NOAA (2018) tropical storm database. Local meteorology data are from the NOAA NCDC (2018) climate portal for three weather stations (Del Rio International Airport TX, Laredo Municipal Airport TX, and Brownsville South Padre Island International Airport TX). Details on these stations and data are in the SI (Methods and Data, additional details). Data Pre-treatment: Global climate indicators, local meteorology, and CFT variables are assembled into a single MS Excel matrix. To address the low signal-to-noise ratio and non-independence of time series common in weather data (SI Methods and Data, additional details, tests). We transform all predictor and response variables using a series of five consecutive steps: 1) first differences (year n minus year n-1) were calculated; 2) and these converted to z scores (z = (x- µ) / s); 3) linear regression was used to remove directional trends; 4) moving averages were calculated for each data vector, and; 5) a lag was optionally applied. The transformed data variables were then tested for predictive ability using simple correlation, probability of, error and level of significance. Bivariate and Multivariate Regression Analysis: Four bivariate Best Model regressions of climate predictors on CFT are developed using XLSTAT software (Addinsoft Inc. 2018); three multivariate models include regression with no interactions, with level 2 interactions, and with variables restricted to two and to four variables minimum. To validate each model, we withhold the first and last 29 observations points. Nine model evaluation and three summary statistics are identified in SI (Methods and Data additional details, definitions).