Efficacy of extracting indices from large-scale acoustic recordings to monitor biodiversity

被引:133
|
作者
Buxton, Rachel T. [1 ]
McKenna, Megan F. [2 ]
Clapp, Mary [3 ]
Meyer, Erik [4 ]
Stabenau, Erik [5 ]
Angeloni, Lisa M. [6 ]
Crooks, Kevin [1 ]
Wittemyer, George [1 ]
机构
[1] Colorado State Univ, Dept Fish Wildlife & Conservat Biol, Ft Collins, CO 80523 USA
[2] Natl Pk Serv, Nat Sounds & Night Skies Div, Ft Collins, CO USA
[3] Univ Calif Davis, Evolut & Ecol Dept, Davis, CA 95616 USA
[4] Sequoia & Kings Canyon Natl Pk, Three Rivers, CA USA
[5] South Florida Nat Resources Ctr, Everglades Natl Pk, Homestead, FL USA
[6] Colorado State Univ, Dept Biol, Ft Collins, CO 80523 USA
关键词
acoustic indices; bioacoustics; biodiversity; passive acoustic monitoring; random forest; LANDSCAPE; SOUNDSCAPE; DIVERSITY; IMPROVE;
D O I
10.1111/cobi.13119
中图分类号
X176 [生物多样性保护];
学科分类号
090705 ;
摘要
Passive acoustic monitoring could be a powerful way to assess biodiversity across large spatial and temporal scales. However, extracting meaningful information from recordings can be prohibitively time consuming. Acoustic indices (i.e., a mathematical summary of acoustic energy) offer a relatively rapid method for processing acoustic data and are increasingly used to characterize biological communities. We examined the relationship between acoustic indices and the diversity and abundance of biological sounds in recordings. We reviewed the acoustic-index literature and found that over 60 indices have been applied to a range of objectives with varying success. We used 36 of the most indicative indices to develop a predictive model of the diversity of animal sounds in recordings. Acoustic data were collected at 43 sites in temperate terrestrial and tropical marine habitats across the continental United States. For terrestrial recordings, random-forest models with a suite of acoustic indices as covariates predicted Shannon diversity, richness, and total number of biological sounds with high accuracy (R(2)0.94, mean squared error [MSE] 170.2). Among the indices assessed, roughness, acoustic activity, and acoustic richness contributed most to the predictive ability of models. Performance of index models was negatively affected by insect, weather, and anthropogenic sounds. For marine recordings, random-forest models poorly predicted Shannon diversity, richness, and total number of biological sounds (R-2 0.40, MSE 195). Our results suggest that using a combination of relevant acoustic indices in a flexible model can accurately predict the diversity of biological sounds in temperate terrestrial acoustic recordings. Thus, acoustic approaches could be an important contribution to biodiversity monitoring in some habitats. Eficiencia de la Extraccion de indices a partir de Registros Acusticos a Gran Escala para Monitorear la Biodiversidad El monitoreo acustico pasivo podria ser una manera poderosa de evaluar la biodiversidad en escalas temporales y espaciales grandes. Sin embargo, la extraccion de informacion significativa a partir de grabaciones puede ser inasequible y requerir de mucho tiempo. Los indices acusticos (es decir, un resumen matematico de la energia acustica) proporcionan un metodo relativamente rapido para procesar los datos acusticos y cada vez se usan mas para caracterizar las comunidades biologicas. Examinamos la relacion entre los indices acusticos y la diversidad y abundancia de sonidos biologicos en las grabaciones. Revisamos la bibliografia sobre el indice de acustica y encontramos que mas de 60 indices han sido aplicados a una gama de objetivos con exito variante. Usamos 36 de los indices mas indicativos para desarrollar un modelo predictivo de la diversidad de sonidos de animales en las grabaciones. Se recolectaron datos acusticos en 43 sitios en habitats terrestres templados y marinos tropicales en todos los Estados Unidos continentales. Para las grabaciones terrestres, los modelos de bosques aleatorios junto con un juego de indices acusticos como covariantes predijeron la diversidad de Shannon, la riqueza y el numero total de sonidos biologicos con una certeza elevada (R-2 0.94, error medio al cuadrado [MSE] 170.2). Entre los indices que se evaluaron, la desigualdad, la actividad acustica y la riqueza acustica fueron los que mas contribuyeron a la habilidad predictiva de los modelos. El desempeno de los modelos de indices fue afectado negativamente por sonidos de insectos, del clima y de origen humano. Para las grabaciones marinas, los modelos de bosque aleatorio predijeron pobremente la diversidad de Shannon, la riqueza y el numero total de sonidos biologicos (R-2 0.40, MSE 195). Nuestros resultados sugieren que el uso de una combinacion de indices acusticos relevantes dentro de un modelo flexible puede predecir con exactitud la diversidad de los sonidos biologicos en un registro acustico de un habitat terrestre templado. Asi, las estrategias acusticas podrian ser una contribucion importante para el monitoreo de la biodiversidad en algunos habitats. Resumen ?? ???????????????????????????????, ??????????????????????? ( ?????????) ???????????????????, ????????????????????????????????????????????????????????????, ??????? 60 ???????????, ?????????36 ?????????, ????????????????????????????????????????? 43 ??????????, ????????????????????????????????????????????? (R(2)0.94, ???? [MSE ] 170.2) ?????????, ??????????????????????????????????????????????????????????????, ????????????????????????????? (R-2 0.40, MSE 195) ????????, ???????????????????????????????????????????, ?????????????????????????????: ???; ??: ???? Article impact statement: Acoustic-processing techniques provide indices that predict the degree of biotic diversity in large-scale passive acoustic recordings.
引用
收藏
页码:1174 / 1184
页数:11
相关论文
共 50 条
  • [1] Techniques for extracting single-trial activity patterns from large-scale neural recordings
    Churchland, Mark M.
    Yu, Byron M.
    Sahani, Maneesh
    Shenoy, Krishna V.
    [J]. CURRENT OPINION IN NEUROBIOLOGY, 2007, 17 (05) : 609 - 618
  • [2] Snapshot Safari: A large-scale collaborative to monitor Africa's remarkable biodiversity
    Pardo, Lain E.
    Bombaci, Sara
    Huebner, Sarah E.
    Somers, Michael J.
    Fritz, Herve
    Downs, Colleen
    Guthmann, Abby
    Hetem, Robyn S.
    Keith, Mark
    le Roux, Aliza
    Mgqatsa, Nokubonga
    Packer, Craig
    Palmer, Meredith S.
    Parker, Daniel M.
    Peel, Mike
    Slotow, Rob
    Strauss, W. Maartin
    Swanepoel, Lourens
    Tambling, Craig
    Tsie, Nairobi
    Vermeulen, Mika
    Willi, Marco
    Jachowski, David S.
    Venter, Jan A.
    [J]. SOUTH AFRICAN JOURNAL OF SCIENCE, 2021, 117 (1-2)
  • [3] Extracting large-scale knowledge bases from the web
    Kumar, R
    Raghavan, P
    Rajagopalan, S
    Tomkins, A
    [J]. PROCEEDINGS OF THE TWENTY-FIFTH INTERNATIONAL CONFERENCE ON VERY LARGE DATA BASES, 1999, : 639 - 650
  • [4] LARGE-SCALE HANDEL + RECENT RECORDINGS
    SADIE, S
    [J]. MUSICAL TIMES, 1986, 127 (1722): : 444 - 444
  • [5] Extracting low-dimensional dynamics from multiple large-scale neural population recordings by learning to predict correlations
    Nonnenmacher, Marcel
    Turaga, Srinivas C.
    Macke, Jakob H.
    [J]. ADVANCES IN NEURAL INFORMATION PROCESSING SYSTEMS 30 (NIPS 2017), 2017, 30
  • [6] Extracting Temporal Patterns from Large-Scale Text Corpus
    Liu, Yu
    Hua, Wen
    Zhou, Xiaofang
    [J]. DATABASES THEORY AND APPLICATIONS (ADC 2019), 2019, 11393 : 17 - 30
  • [7] Biodiversity in groundwater: a large-scale view
    Danielopol, DL
    Pospisil, P
    Rouch, R
    [J]. TRENDS IN ECOLOGY & EVOLUTION, 2000, 15 (06) : 223 - 224
  • [8] Monitor Placement for Large-Scale Systems
    Talele, Nirupama
    Teutsch, Jason
    Erbacher, Robert
    Jaeger, Trent
    [J]. PROCEEDINGS OF THE 19TH ACM SYMPOSIUM ON ACCESS CONTROL MODELS AND TECHNOLOGIES (SACMAT'14), 2014, : 29 - 40
  • [9] Spatial information in large-scale neural recordings
    Cybulski, Thaddeus R.
    Glaser, Joshua I.
    Marblestone, Adam H.
    Zamft, Bradley M.
    Boyden, Edward S.
    Church, George M.
    Kording, Konrad P.
    [J]. FRONTIERS IN COMPUTATIONAL NEUROSCIENCE, 2015, 8 : 1 - 16
  • [10] Dimensionality reduction for large-scale neural recordings
    Cunningham, John P.
    Yu, Byron M.
    [J]. NATURE NEUROSCIENCE, 2014, 17 (11) : 1500 - 1509