REVIEW Open Access

A review of fire effects across SouthAmerican ecosystems the role of climateand time since fireMelisa A Giorgis12 Sebastian R Zeballos1 Lucas Carbone1 Heike Zimmermann3 Henrik von Wehrden4Ramiro Aguilar1 Ana E Ferreras1 Paula A Tecco12 Esteban Kowaljow12 Fernando Barri5 Diego E Gurvich12Pablo Villagra67 and Pedro Jaureguiberry1

Abstract

Background Fire is an important driver of ecosystem dynamics worldwide However knowledge on broad-scalepatterns of ecosystem and organism responses to fires is still scarce Through a systematic quantitative review ofavailable studies across South America we assessed fire effects on biodiversity and abundance of differentorganisms (ie plants fungi invertebrates and vertebrates) plant fitness and soil properties under four climatetypes and time since the last fire (ie early and late post fire) We addressed (1) What fire effects have been studiedacross South America (2) What are the overall responses of biodiversity abundance fitness and soil properties tofires (3) How do climate and time since fire modulate those responses

Results We analyzed 160 articles reporting 1465 fire responses on paired burned and unburned conditions We foundno effect of fire on biodiversity or on invertebrate abundance a negative effect on woody plant species and vertebrateabundance and an increase in shrub fitness Soil in burned areas had higher bulk density and pH and lower organicmatter and nitrogen Fire effect was significantly more positive at early than at late post fire for plant fitness and for soilphosphorus and available nitrogen Stronger negative effects in semiarid climate compared to humid warm climatesuggest that higher temperatures and water availability allow a faster ecosystem recovery after fire

Conclusions Our review highlights the complexity of the climatendashfirendashvegetation feedback when assessing theresponse of soil properties and different organisms at various levels The resilience observed in biodiversity may beexpected considering the large number of fire-prone ecosystems in South America The recovery of invertebrateabundance the reduction of the vertebrate abundance and the loss of nitrogen and organic matter coincide with theresponses found in global reviews at early post-fire times The strength of these responses was further influenced byclimate type and post-fire time Our synthesis provides the first broad-scale diagnosis of fire effects in South Americahelping to visualize strengths weaknesses and gaps in fire research It also brings much needed information fordeveloping adequate land management in a continent where fire plays a prominent socio-ecological role

Keywords abundance biodiversity biomass climate effect size fire impact fire response fitness meta-analysissoil properties

copy The Author(s) 2021 corrected publication 2021 Open Access This article is licensed under a Creative Commons Attribution40 International License which permits use sharing adaptation distribution and reproduction in any medium or format aslong as you give appropriate credit to the original author(s) and the source provide a link to the Creative Commons licenceand indicate if changes were made The images or other third party material in this article are included in the articles CreativeCommons licence unless indicated otherwise in a credit line to the material If material is not included in the articles CreativeCommons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use you will needto obtain permission directly from the copyright holder To view a copy of this licence visit httpcreativecommonsorglicensesby40

Correspondence mgiorgisimbivunceduar1Instituto Multidisciplinario de Biologia Vegetal (CONICET-UNC) VelezSarsfield 1611 CC 495 5000 Cordoba Argentina2Facultad de Ciencias Exactas Fisicas y Naturales Universidad Nacional deCordoba Velez Sarsfield 299 CC 495 5000 Cordoba ArgentinaFull list of author information is available at the end of the article

Fire EcologyGiorgis et al Fire Ecology (2021) 1711 httpsdoiorg101186s42408-021-00100-9

Resumen

Antecedentes El fuego es un importante modulador de la dinaacutemica de los ecosistemas en el mundo Sinembargo el conocimiento a gran escala de los patrones de respuesta de ecosistemas y organismos es auacuten escasoA traveacutes de una revisioacuten sistemaacutetica cuantitativa de los estudios disponibles a lo largo de Sudameacuterica abordamoslos efectos del fuego sobre la biodiversidad abundancia de diferentes organismos (ie plantas hongosinvertebrados y vertebrados) aptitud de las plantas y propiedades del suelo bajo diferentes condiciones climaacuteticas(ie cuatro tipos climaacuteticos) y tiempo transcurrido desde el uacuteltimo incendio (ie tiempo post-fuego temprano ytardiacuteo) Respondimos (1) iquestQueacute efectos del fuego se han estudiado en Sudameacuterica (2) iquestCuaacuteles son las respuestasgenerales de la biodiversidad abundancia aptitud y propiedades del suelo al fuego (3) iquestCoacutemo el clima y el tiempopost-fuego modulan esas respuestas

Resultados Analizamos 160 artiacuteculos que reportaron 1465 respuestas al fuego de sitios apareados quemados y noquemados No encontramos efectos del fuego sobre la biodiversidad y abundancia de invertebrados encontramosefectos negativos sobre la abundancia de especies de plantas lentildeosas y de vertebrados y un incremento en laaptitud de los arbustos El suelo en sitios quemados tuvo mayor densidad aparente y pH y menor materia orgaacutenicay Nitroacutegeno El efecto del fuego fue significativamente maacutes positivo en tiempos post-fuego tempranos que entardiacuteos para aptitud de las plantas y para el Foacutesforo y Nitroacutegeno disponible del suelo Ademaacutes encontramosefectos negativos maacutes fuertes en climas semiaacuteridos que en climas huacutemedos caacutelidos sugiriendo que altastemperaturas y disponibilidad de agua permiten una mayor recuperacioacuten de los ecosistemas despueacutes del fuego

Conclusiones Nuestro estudio destaca la complejidad de las relaciones climandashfuegondashvegetacioacuten para abordar lasrespuestas de las propiedades del suelo y de diferentes organismos a varios niveles La resiliencia observada en labiodiversidad puede ser esperada dado el alto nuacutemero de ecosistemas propensos al fuego en Sudameacuterica Larecuperacioacuten de la abundancia de invertebrados la reduccioacuten de la abundancia de vertebrados y las peacuterdidas deNitroacutegeno y Materia Orgaacutenica coincide con las respuestas encontradas en revisiones globales en tiempos cortosdespueacutes del fuego La tendencia de estas respuestas estuvo ademaacutes influenciada por el tipo de clima y tiempo postfuego Nuestra siacutentesis provee el primer diagnoacutestico a gran escala de los efectos del fuego en Sudameacutericaayudando a visualizar las fortalezas debilidades y vaciacuteos en los estudios del fuego Tambieacuten brinda informacioacutenmuy necesaria para el desarrollo de estrategias adecuadas de manejo en un continente en el cual el fuego juegaun papel socio-ecoloacutegico preponderante

IntroductionFire is an important driver in the dynamics of terrestrialecosystems influencing many attributes functions andprocesses (Bowman et al 2009 Staver et al 2011 Archibaldet al 2018) Fire triggers ecosystem succession causingtraceable changes in attributes such as biodiversity at differ-ent trophic levels (Pausas and Keeley 2009 He et al 2019)abundance (Pausas and Ribeiro 2017 Carbone et al 2019)fitness (Garciacutea et al 2016 Carbone and Aguilar 2017) aswell as soil properties (Certini 2005 Pellegrini et al 2018)Understanding these changes imposed by fire ischallenging as there are several factors involved thatact at different spatial and temporal scales (Bowmanet al 2009 Archibald et al 2013 Harris et al 2016Pausas and Dantas 2017)Feedbacks between climate vegetation and fire often

involve complex non-linear and context-dependent rela-tionships influencing ecological processes and evolution-ary aspects of the biota which in turn influence ecosystemresponses to fire (Pausas and Paula 2012 Archibald et al2013 2018 Miller et al 2013 Pausas and Ribeiro 20132017) Climate which defines gradients of temperature

and moisture is a strong regulator of ecosystem propertiesand plant traits such as biomass leaf area and net pri-mary productivity (Whittaker and Marks 1975 Bond et al2005 Pausas and Ribeiro 2013) Fire activity is linked tothe productivityndasharidity gradient imposed by climate inwhich conditions at either end of the gradient are lessconducive to fire either due to excess of moisture (high-productivity extreme) or lack of fuel (high-aridity extreme)(Pausas and Bradstock 2007 Krawchuk and Moritz 2011Pausas and Ribeiro 2013) Moreover this relationship maybe modulated by characteristics of the vegetation makingthe flammability threshold of a given ecosystem contextdependent (Cochrane 2003 Archibald et al 2009 Parisienand Moritz 2009 Bradstock 2010 Pausas and Paula 2012)However few studies have been able to broaden thespatial scale (eg from regional to continental) to allow usto understand the potential of the climatendashvegetationndashfiresystem dynamics as a modulator of ecosystem response tofire (Archibald et al 2013 2018 Lehmann et al 2014Pausas and Ribeiro 2017)Fire-prone ecosystems many of which occupy the

intermediate range of the productivityndasharidity gradient

Giorgis et al Fire Ecology (2021) 1711 Page 2 of 20

feature flammable seasonal forests grasslands andshrublands (Bond et al 2005) where most of the specieshave life-history traits that allow them to cope with fire(Bond and Keeley 2005 Keeley et al 2011 Bond andVan Wilgen 2012 Pausas and Keeley 2014) In theseenvironments fire is expected to have a mild impactand ecosystem properties are expected to return to thepre-fire state in the short term On the contrary in non-fire-prone ecosystems where climate maintains a highlevel of moisture throughout the year (eg tropical for-est) or limits biomass productivity generating a patchyconfiguration of vegetation (eg desert) most species donot have life-history traits to cope with fires or they areless efficient at coping than species in fire-prone ecosys-tems Consequently even though climatic conditions insome of these regions might lead to high productivityand hence expected rapid biomass recovery (eg humidtropical regions) fires usually have higher and longer-lasting negative impacts on their ecosystem properties(Gerwing 2002 Cochrane 2003 Barlow and Peres 2008Mestre et al 2013 Silveira et al 2016) Moreover innon-fire-prone systems in which climate restrictsproductivity and hence biomass recovery either by lowtemperature or precipitation a slow recovery after firemight be expected (Casady and Marsh 2010 Nelsonet al 2014) Therefore the effect of fire at a broad scalemeasured through biophysical variables such as biodiver-sity abundance fitness (eg vegetative reproductive)and soil properties (eg bulk density organic matter nu-trients) is expected to differ across regions with differentclimates (Moretti et al 2009 Miller et al 2013 Lehmannet al 2014)Fire is a discrete disturbance in time and space thus

as post-fire succession advances different responses mayoccur in the same ecosystem (Chapin III et al 2011Mestre et al 2013 Cohn et al 2015) For example post-fire recovery of woody ecosystems has resulted inincreased plant diversity soon after fire due to anincrease in grasses and annual plant diversity followedby a reduction in biodiversity as woody cover increasesover time (Giorgis et al 2013 Maestre et al 2016Doherty et al 2017) Similarly the release of nutrientsthe availability of light and the temporary decrease ofcompetition usually associated with the post-fire envi-ronments can maximize plant fitness of fast-growingand small-sized species (eg grasses and forbs) shortlyafter fire which then decline as the forest canopy re-covers (Keeley et al 2005 Rostagno et al 2006 ChapinIII et al 2011 Kunst et al 2015 Pilon et al 2021)Along the same lines increased soil nutrient pulses thatoccur immediately after fire might drop after some timedue to soil erosion (Certini 2005) Furthermore abun-dance of certain animals (ie invertebrates or verte-brates) could increase shortly after fire as some species

take refuge during fire and then recolonize recentlyburned patches covered by grasses and forbs whileothers could be positively associated with patches inlater stages of post-fire succession where woody andunderstory plants have recovered (Robinson et al 2013Doherty et al 2017 Pausas 2019) Therefore fires mightgenerate windows of opportunity for growth duringwhich different species are at an optimum at differenttimes after fire occurrence (Miller et al 2013 Farns-worth et al 2014 Cohn et al 2015 Pausas and Ribeiro2017) making time since fire a relevant variable forassessing post-fire dynamics (Whelan et al 2002 Mur-phy and Russell-Smith 2010 McLauchlan et al 2020)Since most studies are limited in time a comparativesynthetic approach of individual studies conducted atdifferent times after fire occurrence would allow asses-sing changes over timeSouth America holds one of the highest annual

average numbers of fires worldwide (Andela et al 2017)Most fires on this continent have an anthropogenicorigin (Bowman et al 2009) and occur in the tropics(mainly confined to northeast Brazil Dwyer et al 2000)during late winter and spring (Di Bella et al 2006)Given the great climatic variability within this continent(ie from arid to humid and from cold to warm) weexpect to find important different fire responses acrossecosystems and organisms Although fires are highlyrelevant for the dynamics of South American ecosys-tems knowledge of their ecological impact on ecosystemattributes and processes is scarce with respect to otherfire-prone areas of the world (Prichard et al 2017 Gearyet al 2019) Here we present the first systematic reviewof fire responses of ecosystem and organism attributesacross South America Our study aims to providegeneral patterns of fire effects at the continental scaleand improve our understanding of the main factorscontrolling post-disturbance dynamics at this scale (ieclimate and time since fire) To that end we performeda systematic review and hierarchical meta-analysis of theexisting literature of fire studies in South America to ad-dress three major questions (1) What fire effects havebeen studied across South America (2) What are theoverall responses of biodiversity abundance fitness andsoil properties to fires (3) How does climate and timesince fire modulate those responses Finally we givesome recommendations for future research that mayhelp improve our understanding of fire effects in SouthAmerica and the world

MethodsLiterature search and compilation of datasetWe searched for peer-reviewed publications of fireeffects in South America in Scopus scientific literaturedatabase Relevant search terms were combined into one

Giorgis et al Fire Ecology (2021) 1711 Page 3 of 20

search string which we used to select the titles abstractsand keywords of all articles in relevant areas namelyagricultural and biological sciences environmentalsciences and earth and planetary sciences We usedthe following search string

This search delivered 4535 studies (10 December 2019)of which we assessed their titles and abstracts retaining435 for a further detailed full-text check from which wekept 160 studies These studies were analyzed in full andall relevant data was extracted for the subsequent meta-analysis We included articles in English Spanish andPortuguese We only included articles that provided nu-merical parameters of the response variables of interest tocompute the effect sizes the common metric to conduct ameta-analysis in paired burned (ie treatment) and un-burned (ie control) plots the latter with at least 20 yearssince the last fire and that reported the time since the lastfire We did not include studies that used experimentalheat treatments or studies with confounding factors suchas slash and burn practices or with obvious differences inconditions (eg topographic position grazing pressure)between plots (Additional file 1 Figure 11)

Response variables considered in the analysesWe analyzed the impact of fire on four response variablesbiodiversity abundance fitness and soil propertiesBiodiversity and abundance were analyzed according totwo levels of disaggregation The first level includedplants fungi invertebrates and vertebrates while the sec-ond level included growth forms for plants (forbs grassesshrubs and trees) taxonomic order for invertebrates (eg

Araneae Coleoptera Collembola Formicidae Orthop-tera) and taxonomic class for vertebrates (birds mam-mals reptiles amphibians) Biodiversity was mostlymeasured as richness (number of species) and only a fewstudies used other diversity indices Abundance was mainlyreported as cover percentage biomass (for plant species)or individual number (for trees and animals) Fitness wasconsidered as any trait of the individuals directly or indir-ectly related to their performance (Violle et al 2007) As al-most all of the studies assessing fire effects on fitness wereconducted on plants (98) we analyzed their vegetative(eg specific leaf area growth rate) and reproductive (egnumber of flowers per branch fruit-set) fitness (Violle et al2007) Finally we analyzed each soil property independently(bulk density litter organic matter available nitrogen totalnitrogen phosphorous micronutrients pH and salinity)For simplicity all the variables subjected to analysis of fireresponses will hereafter be referred to as response variablesTo analyze how climate modulates the effect of fire on

response variables we classified each study site based onthe relation between precipitation and temperatureproposed by Whittaker (1975) for the ordination ofworld ecosystems To do that we first obtained themean annual precipitation and the mean annualtemperature of each study site from Worldclim (wwwworldclimorg Hijmans et al 2005) The location of eachstudy site was geo-referenced in decimal degrees anddata were extracted using the extract function fromthe raster packages in R version 340 (R Core Team2019) from bioclimatic data at a scale of 25 arcmin Ifan article had several study sites (with pairs of burnedand unburned plots) separated by more than 15 km weextracted different bioclimatic data for each pair of sitesand thus we obtained data for 175 study sites from the160 selected articles Secondly based on the formulaproposed by Bond et al (2005) which follows the pre-cipitationndashtemperature plane of Whittaker (1975) to de-fine uncertain ecosystems we classified each study siteinto one of the following four climate types semiariddry cold-humid and warm-humid climates Values ofMAP (mean annual precipitation) and MAT (mean an-nual temperature) were used in combination with unitlessconstant values to define climate types (Table 1)Humid climate was further divided into two typesmdash

humid-cold and humid-warmmdashdepending on whether

Table 1 Formulas used to classify each study site from studies that were published between 1990 and 2019 that were used in thereview of fire effects across South American ecosystems to determine the role of climate and time since fire into different climatetypes MAP = Mean Annual Precipitation MAT = Mean Annual Temperature

Climate types Formulas

Humid MAP gt 7143 MAT + 286 and MAP gt ndash1469 MAT2 + 81665 MAT + 475

Semiarid MAP gt 7143 MAT + 286 and MAP lt ndash1469 MAT2 + 81665 MAT + 475

Dry MAP lt 7143 MAT + 286 and MAP lt ndash1469 MAT2 + 81665 MAT + 475

Giorgis et al Fire Ecology (2021) 1711 Page 4 of 20

feature flammable seasonal forests grasslands andshrublands (Bond et al 2005) where most of the specieshave life-history traits that allow them to cope with fire(Bond and Keeley 2005 Keeley et al 2011 Bond andVan Wilgen 2012 Pausas and Keeley 2014) In theseenvironments fire is expected to have a mild impactand ecosystem properties are expected to return to thepre-fire state in the short term On the contrary in non-fire-prone ecosystems where climate maintains a highlevel of moisture throughout the year (eg tropical for-est) or limits biomass productivity generating a patchyconfiguration of vegetation (eg desert) most species donot have life-history traits to cope with fires or they areless efficient at coping than species in fire-prone ecosys-tems Consequently even though climatic conditions insome of these regions might lead to high productivityand hence expected rapid biomass recovery (eg humidtropical regions) fires usually have higher and longer-lasting negative impacts on their ecosystem properties(Gerwing 2002 Cochrane 2003 Barlow and Peres 2008Mestre et al 2013 Silveira et al 2016) Moreover innon-fire-prone systems in which climate restrictsproductivity and hence biomass recovery either by lowtemperature or precipitation a slow recovery after firemight be expected (Casady and Marsh 2010 Nelsonet al 2014) Therefore the effect of fire at a broad scalemeasured through biophysical variables such as biodiver-sity abundance fitness (eg vegetative reproductive)and soil properties (eg bulk density organic matter nu-trients) is expected to differ across regions with differentclimates (Moretti et al 2009 Miller et al 2013 Lehmannet al 2014)Fire is a discrete disturbance in time and space thus

as post-fire succession advances different responses mayoccur in the same ecosystem (Chapin III et al 2011Mestre et al 2013 Cohn et al 2015) For example post-fire recovery of woody ecosystems has resulted inincreased plant diversity soon after fire due to anincrease in grasses and annual plant diversity followedby a reduction in biodiversity as woody cover increasesover time (Giorgis et al 2013 Maestre et al 2016Doherty et al 2017) Similarly the release of nutrientsthe availability of light and the temporary decrease ofcompetition usually associated with the post-fire envi-ronments can maximize plant fitness of fast-growingand small-sized species (eg grasses and forbs) shortlyafter fire which then decline as the forest canopy re-covers (Keeley et al 2005 Rostagno et al 2006 ChapinIII et al 2011 Kunst et al 2015 Pilon et al 2021)Along the same lines increased soil nutrient pulses thatoccur immediately after fire might drop after some timedue to soil erosion (Certini 2005) Furthermore abun-dance of certain animals (ie invertebrates or verte-brates) could increase shortly after fire as some species

take refuge during fire and then recolonize recentlyburned patches covered by grasses and forbs whileothers could be positively associated with patches inlater stages of post-fire succession where woody andunderstory plants have recovered (Robinson et al 2013Doherty et al 2017 Pausas 2019) Therefore fires mightgenerate windows of opportunity for growth duringwhich different species are at an optimum at differenttimes after fire occurrence (Miller et al 2013 Farns-worth et al 2014 Cohn et al 2015 Pausas and Ribeiro2017) making time since fire a relevant variable forassessing post-fire dynamics (Whelan et al 2002 Mur-phy and Russell-Smith 2010 McLauchlan et al 2020)Since most studies are limited in time a comparativesynthetic approach of individual studies conducted atdifferent times after fire occurrence would allow asses-sing changes over timeSouth America holds one of the highest annual

average numbers of fires worldwide (Andela et al 2017)Most fires on this continent have an anthropogenicorigin (Bowman et al 2009) and occur in the tropics(mainly confined to northeast Brazil Dwyer et al 2000)during late winter and spring (Di Bella et al 2006)Given the great climatic variability within this continent(ie from arid to humid and from cold to warm) weexpect to find important different fire responses acrossecosystems and organisms Although fires are highlyrelevant for the dynamics of South American ecosys-tems knowledge of their ecological impact on ecosystemattributes and processes is scarce with respect to otherfire-prone areas of the world (Prichard et al 2017 Gearyet al 2019) Here we present the first systematic reviewof fire responses of ecosystem and organism attributesacross South America Our study aims to providegeneral patterns of fire effects at the continental scaleand improve our understanding of the main factorscontrolling post-disturbance dynamics at this scale (ieclimate and time since fire) To that end we performeda systematic review and hierarchical meta-analysis of theexisting literature of fire studies in South America to ad-dress three major questions (1) What fire effects havebeen studied across South America (2) What are theoverall responses of biodiversity abundance fitness andsoil properties to fires (3) How does climate and timesince fire modulate those responses Finally we givesome recommendations for future research that mayhelp improve our understanding of fire effects in SouthAmerica and the world

MethodsLiterature search and compilation of datasetWe searched for peer-reviewed publications of fireeffects in South America in Scopus scientific literaturedatabase Relevant search terms were combined into one

Giorgis et al Fire Ecology (2021) 1711 Page 3 of 20

search string which we used to select the titles abstractsand keywords of all articles in relevant areas namelyagricultural and biological sciences environmentalsciences and earth and planetary sciences We usedthe following search string

This search delivered 4535 studies (10 December 2019)of which we assessed their titles and abstracts retaining435 for a further detailed full-text check from which wekept 160 studies These studies were analyzed in full andall relevant data was extracted for the subsequent meta-analysis We included articles in English Spanish andPortuguese We only included articles that provided nu-merical parameters of the response variables of interest tocompute the effect sizes the common metric to conduct ameta-analysis in paired burned (ie treatment) and un-burned (ie control) plots the latter with at least 20 yearssince the last fire and that reported the time since the lastfire We did not include studies that used experimentalheat treatments or studies with confounding factors suchas slash and burn practices or with obvious differences inconditions (eg topographic position grazing pressure)between plots (Additional file 1 Figure 11)

Response variables considered in the analysesWe analyzed the impact of fire on four response variablesbiodiversity abundance fitness and soil propertiesBiodiversity and abundance were analyzed according totwo levels of disaggregation The first level includedplants fungi invertebrates and vertebrates while the sec-ond level included growth forms for plants (forbs grassesshrubs and trees) taxonomic order for invertebrates (eg

Araneae Coleoptera Collembola Formicidae Orthop-tera) and taxonomic class for vertebrates (birds mam-mals reptiles amphibians) Biodiversity was mostlymeasured as richness (number of species) and only a fewstudies used other diversity indices Abundance was mainlyreported as cover percentage biomass (for plant species)or individual number (for trees and animals) Fitness wasconsidered as any trait of the individuals directly or indir-ectly related to their performance (Violle et al 2007) As al-most all of the studies assessing fire effects on fitness wereconducted on plants (98) we analyzed their vegetative(eg specific leaf area growth rate) and reproductive (egnumber of flowers per branch fruit-set) fitness (Violle et al2007) Finally we analyzed each soil property independently(bulk density litter organic matter available nitrogen totalnitrogen phosphorous micronutrients pH and salinity)For simplicity all the variables subjected to analysis of fireresponses will hereafter be referred to as response variablesTo analyze how climate modulates the effect of fire on

response variables we classified each study site based onthe relation between precipitation and temperatureproposed by Whittaker (1975) for the ordination ofworld ecosystems To do that we first obtained themean annual precipitation and the mean annualtemperature of each study site from Worldclim (wwwworldclimorg Hijmans et al 2005) The location of eachstudy site was geo-referenced in decimal degrees anddata were extracted using the extract function fromthe raster packages in R version 340 (R Core Team2019) from bioclimatic data at a scale of 25 arcmin Ifan article had several study sites (with pairs of burnedand unburned plots) separated by more than 15 km weextracted different bioclimatic data for each pair of sitesand thus we obtained data for 175 study sites from the160 selected articles Secondly based on the formulaproposed by Bond et al (2005) which follows the pre-cipitationndashtemperature plane of Whittaker (1975) to de-fine uncertain ecosystems we classified each study siteinto one of the following four climate types semiariddry cold-humid and warm-humid climates Values ofMAP (mean annual precipitation) and MAT (mean an-nual temperature) were used in combination with unitlessconstant values to define climate types (Table 1)Humid climate was further divided into two typesmdash

humid-cold and humid-warmmdashdepending on whether

Table 1 Formulas used to classify each study site from studies that were published between 1990 and 2019 that were used in thereview of fire effects across South American ecosystems to determine the role of climate and time since fire into different climatetypes MAP = Mean Annual Precipitation MAT = Mean Annual Temperature

Climate types Formulas

Humid MAP gt 7143 MAT + 286 and MAP gt ndash1469 MAT2 + 81665 MAT + 475

Semiarid MAP gt 7143 MAT + 286 and MAP lt ndash1469 MAT2 + 81665 MAT + 475

Dry MAP lt 7143 MAT + 286 and MAP lt ndash1469 MAT2 + 81665 MAT + 475

Giorgis et al Fire Ecology (2021) 1711 Page 4 of 20

search string which we used to select the titles abstractsand keywords of all articles in relevant areas namelyagricultural and biological sciences environmentalsciences and earth and planetary sciences We usedthe following search string

This search delivered 4535 studies (10 December 2019)of which we assessed their titles and abstracts retaining435 for a further detailed full-text check from which wekept 160 studies These studies were analyzed in full andall relevant data was extracted for the subsequent meta-analysis We included articles in English Spanish andPortuguese We only included articles that provided nu-merical parameters of the response variables of interest tocompute the effect sizes the common metric to conduct ameta-analysis in paired burned (ie treatment) and un-burned (ie control) plots the latter with at least 20 yearssince the last fire and that reported the time since the lastfire We did not include studies that used experimentalheat treatments or studies with confounding factors suchas slash and burn practices or with obvious differences inconditions (eg topographic position grazing pressure)between plots (Additional file 1 Figure 11)

Response variables considered in the analysesWe analyzed the impact of fire on four response variablesbiodiversity abundance fitness and soil propertiesBiodiversity and abundance were analyzed according totwo levels of disaggregation The first level includedplants fungi invertebrates and vertebrates while the sec-ond level included growth forms for plants (forbs grassesshrubs and trees) taxonomic order for invertebrates (eg

Araneae Coleoptera Collembola Formicidae Orthop-tera) and taxonomic class for vertebrates (birds mam-mals reptiles amphibians) Biodiversity was mostlymeasured as richness (number of species) and only a fewstudies used other diversity indices Abundance was mainlyreported as cover percentage biomass (for plant species)or individual number (for trees and animals) Fitness wasconsidered as any trait of the individuals directly or indir-ectly related to their performance (Violle et al 2007) As al-most all of the studies assessing fire effects on fitness wereconducted on plants (98) we analyzed their vegetative(eg specific leaf area growth rate) and reproductive (egnumber of flowers per branch fruit-set) fitness (Violle et al2007) Finally we analyzed each soil property independently(bulk density litter organic matter available nitrogen totalnitrogen phosphorous micronutrients pH and salinity)For simplicity all the variables subjected to analysis of fireresponses will hereafter be referred to as response variablesTo analyze how climate modulates the effect of fire on

response variables we classified each study site based onthe relation between precipitation and temperatureproposed by Whittaker (1975) for the ordination ofworld ecosystems To do that we first obtained themean annual precipitation and the mean annualtemperature of each study site from Worldclim (wwwworldclimorg Hijmans et al 2005) The location of eachstudy site was geo-referenced in decimal degrees anddata were extracted using the extract function fromthe raster packages in R version 340 (R Core Team2019) from bioclimatic data at a scale of 25 arcmin Ifan article had several study sites (with pairs of burnedand unburned plots) separated by more than 15 km weextracted different bioclimatic data for each pair of sitesand thus we obtained data for 175 study sites from the160 selected articles Secondly based on the formulaproposed by Bond et al (2005) which follows the pre-cipitationndashtemperature plane of Whittaker (1975) to de-fine uncertain ecosystems we classified each study siteinto one of the following four climate types semiariddry cold-humid and warm-humid climates Values ofMAP (mean annual precipitation) and MAT (mean an-nual temperature) were used in combination with unitlessconstant values to define climate types (Table 1)Humid climate was further divided into two typesmdash

humid-cold and humid-warmmdashdepending on whether

Table 1 Formulas used to classify each study site from studies that were published between 1990 and 2019 that were used in thereview of fire effects across South American ecosystems to determine the role of climate and time since fire into different climatetypes MAP = Mean Annual Precipitation MAT = Mean Annual Temperature

Climate types Formulas

Humid MAP gt 7143 MAT + 286 and MAP gt ndash1469 MAT2 + 81665 MAT + 475

Semiarid MAP gt 7143 MAT + 286 and MAP lt ndash1469 MAT2 + 81665 MAT + 475

Dry MAP lt 7143 MAT + 286 and MAP lt ndash1469 MAT2 + 81665 MAT + 475

Giorgis et al Fire Ecology (2021) 1711 Page 4 of 20

the mean annual temperature of the study sites waslower or higher than 15 degC respectively (Additional file 2)Regarding fire proneness there is no standardized and un-equivocal way to classify the analyzed studies as fire-proneor non-fire-prone instead we relied on the above climate-type classifications and discuss fire proneness when it wasrelevantThe time since the last fire event at which the post-fire

response was measured was classified as either early (be-tween 1 to 36 months) or late (more than 37 months) Re-grettably there were not enough studies reporting the firefrequency (19 of the studies) or the intensity or severity(25 of the studies) of the fire so we were unable to in-clude and analyze these two aspects of the fire regime

Meta-analysisWe used Hedgesrsquo d as an estimate of the unbiasedstandardized mean difference (ie the effect size) betweenthe response variables in burned and unburned condi-tions which has the advantage of being unbiased by smallsample size (Gurevitch et al 2001) To calculate Hedgesrsquod we extracted the mean standard deviation and samplesize values of the response variables in burned andunburned conditions from each study These data wereextracted directly from the text tables or from graphsusing Data Thief III software (Tummers 2006 httpwwwdatathieforg) When some data was missing or notavailable we requested it from the corresponding authorsA positive d value indicates that the response variableincreases in the burned plot whereas a negative d valueimplies the opposite a decrease of the response variable inburned conditions Hedgesrsquo d calculations were performedusing the escalc function from metafor package inthe software RWe performed hierarchical mixed effects meta-analyses

using the rmamv function from metafor package in RThese mixed models were used with fixed and randomeffects to account for differences across studies assumingthat they do not share a common mean effect but thatthere is random variation among studies in addition towithin-study sampling variation (Borenstein et al 2009)In a meta-analytical framework when more than oneeffect size is obtained from the same publication it impliespseudoreplication violating the assumption that effectsizes are independent (Gurevitch and Hedges 1999 Tucket al 2014) To control for this potential source of non-independence among effect sizes we included the identityof each study as a random factor to incorporate theirhierarchical dependence when multiple observations(ie effect sizes) were obtained from the same study(Borenstein et al 2010 Nakagawa and Santos 2012see models in Additional file 3 Tables 31 and 32)Separate meta-analyses for each response variable (ie

biodiversity abundance plant fitness and soil properties)

were performed using different categorical moderatorvariables to analyze descriptive patterns (eg taxa groupsplant life form etc) or to assess whether they affect themagnitude of fire impacts (post-fire time climate type)We also tested combinations between these moderatorswhen the sample size allowedFor each response variable we tested the heterogeneity

of effect sizes using Q statistics which are weightedsums of squares tested against a chi-square distribution(Hedges and Olkin 2014) Specifically we examined theP-values of Q-between (Qb) statistics that describe thevariation in effect sizes that can be attributed to differ-ences among categories of each moderator variableEffect sizes were considered significantly different fromzero if their 95 bias-corrected bootstrap confidence in-tervals (CI) did not include zero (Borenstein et al 2009)A common problem of any systematic review is thepotential inclusion of studies that only show significantresults as they may have a greater possibility of beingpublished than those showing non-significant results(ie publication bias) To detect the existence ofpublication bias in our dataset and to estimate howsuch bias if it existed may affect the overall resultswe used statistical (rank correlation tests and ldquotrimand fillrdquo procedures) and numerical (Rosengbergrsquosfail-safe number) methods (Jennions et al 2013)

ResultsWhat fire effects have been studied in South AmericaA total of 160 articles with 1465 fire responses (effectsizes) were included in our review (Additional file 1)Published articles in South America increased exponen-tially over the years (Fig 1) Most of the studies in thearticles were performed in Brazil (74 studies) andArgentina (64) followed by Chile (10) The remainingSouth American countries each had four articles or less(Fig 1) Among all the considered response variablesfire effects on abundance had the highest number ofrecords (544 effect sizes) followed by plant fitness soilproperties and biodiversity (434 354 and 133 effectsizes respectively) Among taxonomic groups plantshad the highest number of records focusing on plantabundance and fitness while the responses reported forthe abundance of invertebrates and vertebrates weremuch less frequent (Fig 2A) Among plant growthforms trees were responsible for most of the abundanceresponses while shrubs had the majority of the fitnessresponses (Fig 2B) Formicidae was the most studiedorder within invertebrates while birds were responsiblefor most of the studies on vertebrates (Additional file 3Table 32) Most of the studies were performed in semi-arid and humid warm climates (Fig 2C) However inhumid-warm climate studies often focused on abundanceresponses while in semiarid climate fitness responses

Giorgis et al Fire Ecology (2021) 1711 Page 5 of 20

were more common (Fig 2C) Early post-fire re-sponses (lt36 months) were evaluated in most of thestudies while less than half of the responses focused onlate post-fire responses (gt36 months Fig 2D) Studies in-cluding soil properties mostly focused on organic matterfollowed by total nitrogen phosphorus and pH while fewregistered effects of other soil characteristics (Fig 2E)Furthermore most of these variables were studied insemiarid areas (Fig 2E) and in early post fire (Fig 2F)

Fire responses and effects of climate and time since fireWe did not find significant effects of fire on overall bio-diversity (Fig 3 Additional file 3 Table 32) Similarlyno significant fire effects were observed for any of thefour different broad taxonomic groups analyzed (plantsfungi invertebrates and vertebrates Fig 3 Additionalfile 3 Table 32) Moreover biodiversity was not affectedby fire across the two climates that allowed for analysis(humid-warm and semiarid) nor was it affected at differ-ent post-fire times or when only plants were considered(Fig 3 Additional file 3 Table 32)Fire had significantly different effects on the abun-

dance of plants invertebrates and vertebrates (Fig 4)Furthermore fire effect on overall abundance wasnegative for three of the four climates analyzed withthe highest negative effect occurring in dry climatesalthough no significant differences between climateswere found (Fig 4) Additionally fire had a similarnegative effect on overall abundance both at early andlate post-fire times (Fig 4)

For plant abundance fire had an overall significantlynegative effect (Fig 4) which showed the same trend atboth post-fire times (Fig 5A) The negative trend per-sisted when the different growth forms were analyzedseparately (Fig 4 Additional file 3 Table 32) Specific-ally this response was significant in trees and shrubswhile the abundance of forbs and grasses had similarnegative but non-significant fire effects (Fig 4) Thisresponse pattern for different plant growth forms wasequal at both post-fire times (Fig 5B) The fire responseamong different taxa of invertebrates showed an overallslight non-significant negative trend although Orthop-tera was the only order that tended to increase in abun-dance in response to fire (Fig 4) Regarding vertebratesoverall abundance decreased in burned conditionsmostly driven by the negative effect on birds whilemammals reptiles and amphibians were not signifi-cantly affected by fire (Fig 4 Additional file 3 Table32) Vertebrate abundance decreased on average inearly post-fire scenarios but showed no effect in late postfire indicating some level of recovery along post-firesuccession although the differences between times werenot statistically significant (Fig 5A) Except for inverte-brates abundance responses of shrubs trees and verte-brates showed similar trends across the analyzed climatetypes (Fig 5C)We found significantly different fire effects on plant

fitness among different plant growth forms and betweenearly and late post-fire times (Fig 6A Additional file 3Table 32) Fire produced an increase on vegetative andreproductive shrub fitness which was particularly

Fig 1 (A) Geographical distribution of study sites reporting fire effects across South America Different colors indicate the climate typesestablished for the collected data Humid-warm (green) Humid-cold (blue) Semiarid (yellow) and Dry (orange) (B) Cumulative number of studiesby country and year of publication selected for the present study (n = 160) The studies used in the review of fire effects across South Americanecosystems to determine the role of climate and time since fire were published between 1990 and 2019

Giorgis et al Fire Ecology (2021) 1711 Page 6 of 20

evident early post fire (Fig 6B C) Grasses showed asignificant negative response to fire in vegetative fitnessonly during early post-fire while fire effect had a positivetrend on reproductive performance (Fig 6B C) Forbsand trees showed an overall non-significant fire effect ontheir fitness (Fig 6A B C) However a trend towardpositive fire effects of the vegetative fitness of trees earlypost fire was observed (Fig 6C)Regarding soil properties in burned we found plots

reductions in litter OM total nitrogen and availablenitrogen and increases in bulk density and pH (Fig 7AAdditional file 3 Table 32) Additionally we found en-richment of phosphorus early post fire which differedsignificantly from late post fire (Fig 7B) The reductionin organic matter and total nitrogen in burned plots wasmainly driven by their response early post fire and insemiarid climates while for available nitrogen it wasmore notable late post fire (Fig 7B) Increasing values in

pH mainly occurred early post fire and in semiaridclimates (Fig 7B Additional file 3 Table 32)

Publication biasWe found no evidence of publication bias for the meta-analyses of fire effects on abundance and biodiversity asobserved by the lack of correlation between sample sizeand effect sizes Furthermore we found no changes inoverall effect sizes in the ldquotrim and fillrdquo procedures(Additional file 4 Table 41) In contrast for the meta-analyses of fire effects on plant fitness and soil proper-ties we found low but significant correlations betweeneffect sizes and sample sizes indicating the potentialpresence of publication bias (Additional file 4 Table41) However the ldquotrim and fillrdquo methods only slightlychanged the overall results for each of these meta-analyses Moreover all of the calculated weighted fail-safe numbers were always larger than the threshold

Fig 2 Number of fire responses (ie effect sizes) for the studied response and moderator variables across South American ecosystems Panels Athrough D show the number of fire responses for biodiversity abundance and fitness across (A) taxa groups (plants fungi invertebrates [Invert]and vertebrates [Verteb]) (B) plant growth forms (forbs grasses shrubs and trees) (C) climate types (humid-warm humid-cold dry semiarid)and (D) post-fire time (early and late) Panels E and F show the number of fire responses across different soil properties bulk density litterorganic matter (OM) total nitrogen (Total N) nitrogen available (N avail) phosphorous (P) micro nutrients pH and salinity for (E) climate typesand (F) post-fire time The studies used in the review of fire effects across South American ecosystems to determine the role of climate and timesince fire were published between 1990 and 2019

Giorgis et al Fire Ecology (2021) 1711 Page 7 of 20

number of 5n+10 (where n is the number of studies in-cluded in the meta-analysis) implying that the resultsfound in our review were robust regardless of the pres-ence of publication bias (Additional file 4 Table 41)

DiscussionThis is the first systematic quantitative review assessingfire effects on multiple taxonomic groups across differ-ent climate types and post-fire times in South Americaone of the worldrsquos most burned continents Our resultsshow non-significant effects of fire on biodiversity asharp negative effect on abundance of woody plant andvertebrate species and an increase in plant fitness ofshrubs In addition we observed clear negative fireeffects in the most important soil properties as waspreviously found at a global scale Our meta-analysisshowed a tendency of stronger negative effect in semi-arid climate than in tropical warm climate indicatingthat higher temperatures and water availability in thelatter climate allow a faster ecosystem recovery after fireAdditionally more significant fire effects were observed

early post fire than late post fire indicating strongereffects immediately after fire and their dilution over time

Effect of fire on the response variablesBiodiversity Biodiversity was not significantly affected byfire across any taxa group climate type or time since fireindicating an overall high resilience at the analyzed spatialand temporal scales This is in line with previous studiesfocused on specific semiarid regions including two well-studied large fire-prone ecosystems in South America (ieCerrado and the Gran Chaco) which had shown post-firerecovery of biodiversity for different taxa groups such asplants birds and insects (eg Kunst et al 2003 2015Kowaljow et al 2019 Durigan et al 2020 Pilon et al2021) However these responses may certainly involvecontext-dependent changes in species composition (Duri-gan et al 2020 McLauchlan et al 2020) which could notbe captured in our quantitative synthesisInterestingly we found a tendency toward higher

invertebrate biodiversity in burned plots compared tocontrols This pattern has been reported in differentareas around the world (eg (Knoechelmann and Morais

Fig 3 Weighted-mean effect sizes and 95 bias-corrected confidence intervals of fire on from studies published between 1990 and 2019 thatwere used in the review of fire effects across South American ecosystems to determine the role of climate and time since fire The effect sizes ofoverall biodiversity taxa groups climate post-fire time and the response of plants in early (plants-early) and late (plants-late) post fire are shownParameters with confidence intervals that do not overlap the vertical dotted line (Hedgersquos d = 0) are considered to have a significant positive ornegative effect Sample sizes for each category are shown in parentheses The size of each black dot is proportional to its weight or contributionto the overall mean calculation

Giorgis et al Fire Ecology (2021) 1711 Page 8 of 20

2008 Uehara-Prado et al 2010 Carbone et al 2019Lazarina et al 2019) and could be explained by the highmobility of many invertebrates (eg arthropods) whichcan take advantage of the usually high availability ofplant biomass in early post-fire succession (Mirandaet al 2002 Hoffmann and Andersen 2003) Contrary toinvertebrates diversity of fungi (ie mycorrhizas) tendedto decrease (Fig 3) probably as a response to the glo-bally general negative effect of fire on soil microbialcommunities (Pressler et al 2019) and soil quality (Cer-tini 2005 Pellegrini et al 2018) However fungi and ver-tebrates are groups that deserve further study given the

spatial and temporal scarcity of fire effects studies onthese organisms Currently fungi research is highly re-stricted geographically (all studies from ArgentinaLongo et al 2011 2014) while vertebrates showed thelowest number of studies even though this group is fun-damental for controlling cascading effects across trophiclevels (Bruno and Cardinale 2008 Cavallero et al 2013Kurten 2013 Bauer and Hoye 2014 Dirzo et al 2014Peacuterez-Meacutendez et al 2016)Abundance As expected fire negatively affected plant

abundance mainly driven by the impact on woody species(ie shrubs and trees) The negative impact on these

Fig 4 Weighted-mean effect sizes and 95 bias-corrected confidence intervals of fire on abundance from studies published between 1990 and2019 that were used in the review of fire effects across South American ecosystems to determine the role of climate and time since fire Theeffect sizes of overall taxa groups within plant growth forms invertebrates vertebrates climate types and post-fire time are shown Parameterswith confidence intervals that do not overlap the vertical dotted line (Hedgersquos d = 0) are considered to have a significant positive or negativeeffect Sample sizes for each category are shown in parentheses The size of each black dot is proportional to its weight or contribution to theoverall mean calculation Asterisks () denote significant difference (Q-between [Qb] statistics) among categories ( = P lt 00001)

Giorgis et al Fire Ecology (2021) 1711 Page 9 of 20

Fig 5 Weighted-mean effect sizes and 95 bias-corrected confidence intervals of fire on abundance from studies published between 1990 and 2019that were used in the review of fire effects across South American ecosystems to determine the role of climate and time since fire The effect sizes ofearly (-early) and late (-late) post-fire time for abundance of (A) plants invertebrates and vertebrates and (B) grasses shrubs and trees (C) The effectsizes of shrubs trees invertebrates and vertebrates across different climate types (-dry -semiarid -humid-warm) Parameters with confidence intervalsthat do not overlap the vertical dotted line (Hedgersquos d = 0) are considered to have a significant positive or negative effect Sample sizes for eachcategory are shown in parentheses The size of each black dot is proportional to its weight or contribution to the overall mean calculation None ofthe effect sizes of the groups within each moderator variable are significantly different (see Additional file 3 Table 32 for heterogeneity tests)

Giorgis et al Fire Ecology (2021) 1711 Page 10 of 20

growth forms early and late post fire (Fig 5B) is expectedconsidering that the relative recovery of pre-fire abun-dance (including cover and biomass) is slower than inherbaceous species The strongest negative effect observedfor shrubs could be mainly due to the fact that most ofthe studies were carried out in dry-climate ecosystemswhere the combination of water stress and fire signifi-cantly limits shrub recovery (Casillo et al 2012 Pratt et al2014 Jacobsen et al 2016)After fire a combination of different processes may

limit woody species recovery First woody species innon-fire-prone ecosystems may experience high mortal-ity rates (Barlow et al 2003) Second the reduction inaboveground biomass produced by fires exposes individ-uals to direct influence of browsing and trampling bylivestock (Coop et al 2010 Blackhall et al 2015 2017Marcora et al 2018 OrsquoConnor et al 2020 Zeballos et al2020) Third recruitment of woody species seedlingsinto fire-promoted grasslands or herbaceous patchesmay be more restricted due to an increase in competi-tion for light water or nutrients (Scholes and Archer1997 Casillo et al 2012) Finally soil erosion driven byfire increases run-off and seed loss events and reducesnutrient availability therefore constraining biomass pro-duction (Silva et al 2013 Balch et al 2015) The lessernegative effect observed in forbs and grasses may be at-tributed to their relative higher growth rate and shorterlifespan as compared to woody species which result in afaster recovery in burned habitats This vegetativeresponse might be linked to the responses observed insome of the mobile organisms (Fig 4) For example theearly recovery of herbaceous vegetation could promoterapid re-colonization by invertebrates such as insectherbivores and pollinators Additionally our resultsindicated different trends of abundance recovery acrosstaxa groups of invertebrates and vertebrates Similarly aprevious review showed different patterns of ldquooptimumrdquoabundance as time since fire increased (Doherty et al2017) The fast recovery of invertebrate abundance earlypost fire was mainly driven by insect studies (Carboneet al 2019) particularly on Orthoptera (Silveira et al2010 Kral et al 2017) In turn the overall negative effectof fire on the abundance of vertebrates was mainlydriven by bird studies which mostly showed a negativeresponse to fire (Fig 5) Although this pattern has beenobserved in other ecosystems (Fontaine and Kennedy2012 Doherty et al 2017 Carbone et al 2019) the re-sponse of birds to fire can be heterogeneous varyingconsiderably between taxonomic and functional groups(Fontaine and Kennedy 2012) therefore generalizationsmay be difficult to establish The pattern observed herecould be due to the fact that the systematically selectedstudies focused on rather intense fires or on bird speciesthat are more susceptible to fire Our findings of some

Fig 6 Weighted-mean effect sizes and 95 bias-corrected confidenceintervals of fire on plant fitness from studies published between 1990and 2019 that were used in the review of fire effects across SouthAmerican ecosystems to determine the role of climate and time sincefire (A) Effect sizes within growth forms and early and late post-firetime (B) Effect sizes of vegetative and reproductive fitness withingrowth forms (C) Effect sizes of vegetative (-veget) and reproductive(-reprod) fitness within growth forms and early (-early) and late (-late)post-fire time Parameters with confidence intervals that do notoverlap the vertical dotted line (Hedgersquos d = 0) are considered to havea significant positive or negative effect Sample sizes for each categoryare shown in parentheses The size of each black dot is proportional toits weight or contribution to the overall mean calculation Asterisks ()denotes a significant difference (Q-between [Qb] statistics) amongcategories ( = P lt 00001 = P lt 005)

Giorgis et al Fire Ecology (2021) 1711 Page 11 of 20

level of recovery of vertebrate abundance late post firecompared to early post fire (Fig 5) could have beendriven by the re-establishment of some bird populations

over time (Fontaine and Kennedy 2012 Doherty et al2017) Additionally the recovery of populations ofspecialist birds may take longer particularly in forest

Fig 7 Weighted-mean effect sizes and 95 bias-corrected confidence intervals of fire on soil properties from studies published between 1990and 2019 that were used in the review of fire effects across South American ecosystems to determine the role of climate and time since fire (A)Effect sizes of bulk density litter organic matter (OM) total nitrogen (Total N) available nitrogen (N) phosphorous (P) micronutrients pH andsalinity are shown (B) Effect sizes of early (-early) and late (-late) post-fire time of organic matter total nitrogen available nitrogen phosphorousand within climate type (-humid-warm -humid-cold -semiarid) of organic matter and total nitrogen Parameters with confidence intervals that donot overlap the vertical dotted line (Hedgersquos d = 0) are considered to have a significant positive or negative effect Sample sizes for each categoryare shown in parentheses The size of each black dot is proportional to its weight or contribution to the overall mean calculation Asterisk ()denotes a significant difference (Q-between [Qb] statistics) among categories ( = P lt 005)

Giorgis et al Fire Ecology (2021) 1711 Page 12 of 20

ecosystems (eg Barlow and Peres 2004 Fontaine andKennedy 2012 Mestre et al 2013 Albanesi et al 2014)although there are exceptions (eg Fontaine andKennedy 2012 Lee 2018 Morales et al 2020) such asthe group of birds known as cavity excavators (iewoodpeckers [Picidae]) that take advantage of both thesoftness of the wood of recently burned trees (ie snags)for making their cavities (Schepps et al 1999 Winklerand Christie 2002 Bond et al 2012 Lorenz et al 2015)and the higher amount of food (insect and other inverte-brates) found in that substrate right after the fire(Murphy and Lehnhausen 1998 Nappi et al 2003)Plant fitness Despite the negative effect of fire on tree

abundance we found no effect of fire on tree fitnessand a slight positive trend on vegetative fitness early postfire (Fig 6A B) Many tree species in the analyzed eco-systems are highly resilient to individual fires mainlythrough resprouting (Hoffmann and Solbrig 2003 Sou-chie et al 2017 Jaureguiberry et al 2020) therefore ifburned plots do not burn again trees might grow andrecover their biomass (Doherty et al 2017 Miller et al2019 Coop et al 2020) This response does not neces-sarily imply a direct adaptation to fire as the fire-adaptivetraits on which it relies could also have arisen in responseto other factors that have had a long history in some ofthe studied regions such as herbivory and seasonaldrought (Keeley et al 2011) which indirectly provide thevegetation with a high resilience to fire On the otherhand the negative tendency of the reproductive fitness ob-served early post fire is expected since trees may take along time to recover their reproductive size (Chapin IIIet al 2011) In contrast to trees shrubs did show an in-crease in vegetative and reproductive fitness early post firewhich differs significantly from that of late post fire(Fig 6C) The early vegetative response reflects the highresprouting capacity of many shrub species across thestudied regions especially in fire-prone regions (Gurvichet al 2005 Bravo et al 2014 Torres et al 2014 Duriganet al 2020 Jaureguiberry et al 2020) Regarding the repro-ductive response the pattern reported here agrees withprevious evidence suggesting that shrubs might reachminimum reproductive sizes soon after fire then slowdown their growth rate and reproduction as successionprogresses (Hoffmann and Moreira 2002 Hoffmann andSolbrig 2003 Galiacutendez et al 2009) In the case of herb-aceous plants surprisingly we did not find a general effectin fitness but a positive trend in reproductive fitness wasobserved mainly early post fire for grasses (Fig 6B C)The lack of a consistent general response pattern in forbscould be explained by the high heterogeneity of functionalgroups included within this growth form (eg annual per-ennial climber fern) that can have different responses tofire (Keeley et al 1981 Bates et al 2014 Heydari et al2016 Arcamone and Jaureguiberry 2018 Vidaller et al

2019) In the case of grasses the tendency of increasing re-productive fitness early post fire (Fig 6C) could be relatedto the positive firendashcolonization feedback generally de-scribed for this growth form (Bond et al 2003 Bond andKeeley 2005 Pausas and Keeley 2009 Pilon et al 2018)Previous studies had observed an increase in reproductivefitness of grasses after fire (Baruch and Bilbao 1999 Ara-uacutejo et al 2013 Pilon et al 2018 Vidaller et al 2019) Onthe contrary vegetative fitness showed a negative impactearly post fire However we cannot determine if the ob-served pattern (ie positive reproductive fitness versusnegative vegetative fitness) was due to intrinsic differencesbetween types of fitness or to differences in the type of cli-mate since all studies on reproductive fitness were lo-cated in humid-warm climate while vegetative fitnesswas mostly studied in semiarid and arid climatesSoil properties The responses of soil properties to fire in

South American ecosystems support the trend suggestedin previous global reviews (Certini 2005 Pellegrini et al2018) Fire significantly reduces the amount of two im-portant soil properties OM and nitrogen content (Pelle-grini et al 2018) On the one hand OM is directlyassociated with carbon concentration in the soil one ofthe most stable carbon reservoirs worldwide thus large-scale losses of OM due to fire could have important impli-cations for global climate change (Lal 2004) On the otherhand nitrogen is often the most limiting soil nutrient forplant growth and its loss directly affects ecosystems by re-ducing net primary productivity (Vitousek and Howarth1991 Pellegrini et al 2018) In addition our results of ahigher negative effect of fire on OM in semiarid cli-mates compared to humid-warm climate is supportedby Pellegrini et al (2018) who found that across sa-vannas and grasslands biannual fire simulation induceshigh carbon losses under drier climates However ithas been recently proposed that a single fire event maybe related to either an increase a decrease or nochange in soil carbon and nitrogen while frequent firesover time produce a consistent decrease of both ele-ments across ecosystems (Pellegrini et al 2020a b)This hypothesis poses a constraint to our analysessince most of the articles did not report fire history fortheir study sites Consequently the heterogeneous ef-fects of single fires on soil carbon and nitrogen may beconfused with those of frequent fires therefore limitingthe identification of clearer patternsFire produces an increase in bulk density as a result

of soil disaggregation due to the decrease or loss ofOM and to the infiltration of ashes into soil micro-pores (Boyer and Miller 1994 Mataix-Solera et al2011) Soil pH also increased in burned plots as aresult of organic acid denaturation due to soil heatingand the accumulation of potassium and sodiumhydroxides and magnesium and calcium carbonates

Giorgis et al Fire Ecology (2021) 1711 Page 13 of 20

(Knicker 2007) The increase in pH is also driven bythe reduction in OM (Certini 2005) Additionallyimmediate phosphorus enrichment after a fire event isthe result of organic pool conversion of phosphorusinto orthophosphate which declines at later post-firetimes as it is also suggested by Certini (2005)Climatendashfirendashvegetation feedbacks Our review sug-

gests that climatic conditions associated with higherrates of biomass accumulation might determine a fasterecosystem recovery in burned sites across South Amer-ica We found a faster recovery in invertebrate abun-dance and lower losses of soil organic matter underclimates with higher water availability and temperaturesthan in semiarid climates Beyond climatic conditions inhistorically fire-prone ecosystems where fire has pre-sumably been an important evolutionary factor (Bondand Keeley 2005) a greater recovery of the biota is ex-pected compared to non-fire-prone ecosystems Howeverour classification of climate types may not necessarily re-flect the fire history of the corresponding regions there-fore making a clear interpretation of the reported patternsdifficult Although semiarid climate sites could be clearlyconsidered as fire-prone systems (Bond 2005) warm-humid climate sites may also include fire-prone ecosys-tems because some of them have seasonal rainfall pat-terns which is a key characteristic in determining the fireproneness of an ecosystem (Nogueira et al 2017 Romanoand Ursino 2020) In fact following the classification ofOlson et al (2001) 40 of the studies from warm-humidareas corresponded to tropical and subtropical grasslandsavanna and shrublands biome (arguably fire prone)while the other 60 corresponded to tropical and sub-tropical moist broadleaf forest biome (arguably non-fireprone) Most of the former were located in the Cerradoregion where Hoffmann and Moreira (2002) found a fas-ter post-fire recovery of woody species compared toneighboring tropical forest areas Thus they concludedthat the combination of fire-prone vegetation presumablywith a long fire history and warm-humid climate mightfavor a fast ecosystem recovery after fire Therefore theobserved results of greater recovery in warm-humid eco-systems types is likely due to a combination of both cli-mate and fire history Likewise most of the studies fromcold-humid areas have a mediterranean-type climate (egnorthwest Patagonia in Argentina Keeley et al 2012)where rainy winters and hot and dry summers make theseregions fire prone Such regions have been historicallysubjected to periodic fires (both natural and anthropo-genic) and therefore many plant species are capable of re-covering after a fire event (Veblen et al 2003 2008Defosseacute et al 2015) Interestingly we found a significantpost-fire decrease in nitrogen content and soil OM undersemiarid climate but not under cold-humid climateThis response pattern may be due to the inherent

differences in ecosystem properties between climatetypes (eg decomposition and growth rates succes-sional patterns and fire regimen among othersMcLauchlan et al 2020) Fire regime might play animportant role as more frequent surface fire in semi-arid ecosystems could lead to a lower recovery of soilproperties compared to cold-humid ecosystems wherecrown fires predominate and fire frequency is lower thanin semiarid ecosystems therefore buffering fire effects onsoil propertiesThe lack of a standardized classification of fire-

prone and non-fire-prone regions as well as the lim-ited available information on the frequency intensityand severity of fires in the research studies includedin this review are certainly limitations that call forprecaution in the interpretation and extrapolation ofthe results reported here Overcoming such limita-tions remains a challenge in fire ecology (Harris et al2016 Kelly et al 2018) We believe that our studyrepresents a step forward in the synthesis of fire-related patterns at a large scale Our meta-analysis aswell as the discussed patterns provides relevant infor-mation toward understanding the feedback betweenclimate vegetation and fire on the South Americancontinent An ultimately desirable goal in the mediumor long term would be the development of a globalnetwork of fire field experimental surveys across sys-tems with different vegetation types and fire histories

Future perspectivesAcross recently published global fire reviews there isclearly a low representation of studies from SouthAmerica (Prichard et al 2017 Geary et al 2019)Given the relevance of this continent in the globaldynamics of fire it is highly necessary that morestudies on different aspects of fire ecology be carriedout throughout fire-prone regions of the continent(eg seasonal forests and humid and semiarid savannas)Furthermore our review allowed us to identify specific re-search gaps which can be added to the agenda of researchpriorities for future studies in fire ecology (McLauchlanet al 2020)Most of the analyzed studies failed to include fire

characteristics such as frequency intensity and severitywhich prevented us from obtaining a reliable pattern oftheir role as modulators of ecosystem and organismresponses to fire This is an unmet challenge in fireresearch as these fire characteristics may certainly influ-ence the trends reported in this review (Keeley 2009Fontaine and Kennedy 2012 Balch et al 2015 Silveiraet al 2016 Carbone et al 2019)Considering that woody ecosystems might take more

than 40 years to recover after fire (Cavallero et al 2015

Giorgis et al Fire Ecology (2021) 1711 Page 14 of 20

Doherty et al 2017) studies on fire responses that spana short recovery period fail to elucidate what might hap-pen in the meantime between fire occurrence and recov-ery Consequently our review poses an urgent call toincrease the studies assessing changes in fire effects overlonger post-fire times than those usually consideredFurther studies focusing on biodiversity are needed to

increase sensibility to detect changes across taxa groupsclimate types and post-fire timesAdditionally there is a gap of studies on vertebrate

responses to fire and some of the few available ones havemethodological limitations which prevented us from in-cluding them in our analysis (eg lack of replicates inmammals studies Griffiths and Brook 2014)Studies dealing with fire and exotic or invasive species

were quite scarce While invasion ecology has shown in-creasing attention in the last decades (eg Gurevitch et al2011 Pyšek et al 2020) it does not seem to be reflected infire studies in South America (but see Chaneton et al2004 Hoffmann et al 2004 Maziacutea et al 2010 Raffaeleet al 2016 Herrero et al 2016 Marcora et al 2018) Webelieve that this should be a priority topic for future re-search particularly considering the tight link betweenplant invasions and changes in fire regimes (Brooks et al2004 Mandle et al 2011 Harris et al 2016)

Supplementary InformationThe online version contains supplementary material available at httpsdoiorg101186s42408-021-00100-9

Additional file 1 Supplementary material regarding the choice ofstudies published between 1990 and 2019 that were used in the reviewof fire effects across South American ecosystems to determine the role ofclimate and time since fire Figure 11 PRISMA flow diagramrepresenting the flow of information through the decision process (iethe number of studies identified rejected and accepted) for the meta-analysis of fire effects across South American ecosystems to determinethe role of climate and time since fire Studies used were published be-tween 1990 and 2019 List 11 List of reviewed studies published be-tween 1990 and 2019 that were used in the meta-analysis of fire effectsacross South American ecosystems to determine the role of climate andtime since fire

Additional file 2 Distribution across the climatic temperaturendashprecipitation plane of study sites (n = 175) from the 160 articles reviewedfor the meta-analysis of fire effects across South American ecosystems todetermine the role of climate and time since fire Studies used were pub-lished between 1990 and 2019 The four climate types established for thecollected data are indicated by different colors

Additional file 3 Distribution across the climatic temperaturendashprecipitation plane of study sites (n = 175) from the 160 articles reviewedfor the meta-analysis of fire effects across South American ecosystems todetermine the role of climate and time since fire Studies used were pub-lished between 1990 and 2019 The four climate types established for thecollected data are indicated by different colors Table 31 Synopsis ofmodels used in the different hierarchical mixed effects meta-analyses offire effects across South American ecosystems taken from studies pub-lished between 1990 and 2019 to determine the role of climate and timesince fire expressed in R language syntax using the metafor package(Viechtbauer 2010) yi = Hedgesrsquo d vi = variance of Hedgesrsquo d ID = effectidentity Table 32 Summary of the fire effects on biodiversity abun-dance fitness and soil properties showing tests of moderator variables

and heterogeneities from the meta-analyses of fire effects across SouthAmerican ecosystems to determine the role of climate and time sincefire Studies used were published between 1990 and 2019 Significant in-fluence of moderators is indicated in bold when P le 005 and the 95confidence interval (CI) does not include zero Non-significant effects (P gt005) are indicated with ldquonsrdquo Note that Q-total (QT) statistics are reportedfor overall effect and Q-between (Qb) statistics are reported for categor-ical moderator variables

Additional file 4 Supplementary information 41 Publication biasfor the meta-analysis of fire effects across South American ecosystems todetermine the role of climate and time since fire Studies used were pub-lished between 1990 and 2019 Table 41 Sample size (number of obser-vations) and results of publication bias and hierarchical models testingfor each response variable in our meta-analysis of fire effects across SouthAmerican ecosystems to determine the role of climate and time sincefire Studies used were published between 1990 and 2019 Publicationbias was tested using fail-safe number (ie Rosenberg method) Kendallrsquosrank correlation test (z) and ldquotrim and fillrdquo models

AcknowledgementsMAG and HvW were partially supported by the National Scientific andTechnological Research Council (CONICET) fellow program MAG SRZ LC RAAF PAT EK FB DEG PV and PJ are researchers of CONICET Argentina MAGLC PAT EK FB and DEG are professors at the National University of Cordobaand PV is professor at the Universidad Nacional de Cuyo

Authorsrsquo contributionsMAG SRZ LC RA HZ HvW and PJ designed the research and analysesMAG SRZ LC RA and PJ conducted literature search with the contributionof all the authors MAG LC and PJ wrote the article with contributions fromRA AF PT and EK All authors read and approved the final manuscript

FundingThe National Council for Scientific and Technological Research (CONICET) ofArgentina indirectly supported this manuscript through exchange grants forMAG and HvW However no direct funding supported this project

Availability of data and materialsNo unique data were created in the writing of this manuscript so noadditional data are available

Declarations

Ethics approval and consent to participateNot applicable

Consent for publicationConsent for publication not applicable as we use previously published

Competing interestsThe authors declare that they have no competing interests

Author details1Instituto Multidisciplinario de Biologia Vegetal (CONICET-UNC) VelezSarsfield 1611 CC 495 5000 Cordoba Argentina 2Facultad de CienciasExactas Fisicas y Naturales Universidad Nacional de Cordoba Velez Sarsfield299 CC 495 5000 Cordoba Argentina 3Faculty of Sustainability LeuphanaUniversity Luneburg Institute for Ethics and Transdisciplinary SustainabilityResearch Universitatsallee 1 21335 Luneburg Germany 4Faculty ofSustainability Leuphana University Luneburg Institute of EcologyUniversitatsallee 1 21335 Luneburg Germany 5Instituto de Diversidad yEcologia Animal (IDEA) CONICET-UNC and Facultad de Ciencias ExactasFisicas y Naturales Universidad Nacional de Cordoba Velez Sarsfield 299 CP5000 Cordoba Argentina 6Facultad de Ciencias Agrarias UniversidadNacional de Cuyo Brown sin nuacutemero Chacras de Coria 5505 MendozaArgentina 7Facultad de Ciencias Agrarias Universidad Nacional de CuyoM5528AHB Chacras de Coria Mendoza Argentina

Giorgis et al Fire Ecology (2021) 1711 Page 15 of 20

evident early post fire (Fig 6B C) Grasses showed asignificant negative response to fire in vegetative fitnessonly during early post-fire while fire effect had a positivetrend on reproductive performance (Fig 6B C) Forbsand trees showed an overall non-significant fire effect ontheir fitness (Fig 6A B C) However a trend towardpositive fire effects of the vegetative fitness of trees earlypost fire was observed (Fig 6C)Regarding soil properties in burned we found plots

reductions in litter OM total nitrogen and availablenitrogen and increases in bulk density and pH (Fig 7AAdditional file 3 Table 32) Additionally we found en-richment of phosphorus early post fire which differedsignificantly from late post fire (Fig 7B) The reductionin organic matter and total nitrogen in burned plots wasmainly driven by their response early post fire and insemiarid climates while for available nitrogen it wasmore notable late post fire (Fig 7B) Increasing values in

pH mainly occurred early post fire and in semiaridclimates (Fig 7B Additional file 3 Table 32)

Publication biasWe found no evidence of publication bias for the meta-analyses of fire effects on abundance and biodiversity asobserved by the lack of correlation between sample sizeand effect sizes Furthermore we found no changes inoverall effect sizes in the ldquotrim and fillrdquo procedures(Additional file 4 Table 41) In contrast for the meta-analyses of fire effects on plant fitness and soil proper-ties we found low but significant correlations betweeneffect sizes and sample sizes indicating the potentialpresence of publication bias (Additional file 4 Table41) However the ldquotrim and fillrdquo methods only slightlychanged the overall results for each of these meta-analyses Moreover all of the calculated weighted fail-safe numbers were always larger than the threshold

Fig 2 Number of fire responses (ie effect sizes) for the studied response and moderator variables across South American ecosystems Panels Athrough D show the number of fire responses for biodiversity abundance and fitness across (A) taxa groups (plants fungi invertebrates [Invert]and vertebrates [Verteb]) (B) plant growth forms (forbs grasses shrubs and trees) (C) climate types (humid-warm humid-cold dry semiarid)and (D) post-fire time (early and late) Panels E and F show the number of fire responses across different soil properties bulk density litterorganic matter (OM) total nitrogen (Total N) nitrogen available (N avail) phosphorous (P) micro nutrients pH and salinity for (E) climate typesand (F) post-fire time The studies used in the review of fire effects across South American ecosystems to determine the role of climate and timesince fire were published between 1990 and 2019

Giorgis et al Fire Ecology (2021) 1711 Page 7 of 20

number of 5n+10 (where n is the number of studies in-cluded in the meta-analysis) implying that the resultsfound in our review were robust regardless of the pres-ence of publication bias (Additional file 4 Table 41)

DiscussionThis is the first systematic quantitative review assessingfire effects on multiple taxonomic groups across differ-ent climate types and post-fire times in South Americaone of the worldrsquos most burned continents Our resultsshow non-significant effects of fire on biodiversity asharp negative effect on abundance of woody plant andvertebrate species and an increase in plant fitness ofshrubs In addition we observed clear negative fireeffects in the most important soil properties as waspreviously found at a global scale Our meta-analysisshowed a tendency of stronger negative effect in semi-arid climate than in tropical warm climate indicatingthat higher temperatures and water availability in thelatter climate allow a faster ecosystem recovery after fireAdditionally more significant fire effects were observed

early post fire than late post fire indicating strongereffects immediately after fire and their dilution over time

Effect of fire on the response variablesBiodiversity Biodiversity was not significantly affected byfire across any taxa group climate type or time since fireindicating an overall high resilience at the analyzed spatialand temporal scales This is in line with previous studiesfocused on specific semiarid regions including two well-studied large fire-prone ecosystems in South America (ieCerrado and the Gran Chaco) which had shown post-firerecovery of biodiversity for different taxa groups such asplants birds and insects (eg Kunst et al 2003 2015Kowaljow et al 2019 Durigan et al 2020 Pilon et al2021) However these responses may certainly involvecontext-dependent changes in species composition (Duri-gan et al 2020 McLauchlan et al 2020) which could notbe captured in our quantitative synthesisInterestingly we found a tendency toward higher

invertebrate biodiversity in burned plots compared tocontrols This pattern has been reported in differentareas around the world (eg (Knoechelmann and Morais

Fig 3 Weighted-mean effect sizes and 95 bias-corrected confidence intervals of fire on from studies published between 1990 and 2019 thatwere used in the review of fire effects across South American ecosystems to determine the role of climate and time since fire The effect sizes ofoverall biodiversity taxa groups climate post-fire time and the response of plants in early (plants-early) and late (plants-late) post fire are shownParameters with confidence intervals that do not overlap the vertical dotted line (Hedgersquos d = 0) are considered to have a significant positive ornegative effect Sample sizes for each category are shown in parentheses The size of each black dot is proportional to its weight or contributionto the overall mean calculation

Giorgis et al Fire Ecology (2021) 1711 Page 8 of 20

2008 Uehara-Prado et al 2010 Carbone et al 2019Lazarina et al 2019) and could be explained by the highmobility of many invertebrates (eg arthropods) whichcan take advantage of the usually high availability ofplant biomass in early post-fire succession (Mirandaet al 2002 Hoffmann and Andersen 2003) Contrary toinvertebrates diversity of fungi (ie mycorrhizas) tendedto decrease (Fig 3) probably as a response to the glo-bally general negative effect of fire on soil microbialcommunities (Pressler et al 2019) and soil quality (Cer-tini 2005 Pellegrini et al 2018) However fungi and ver-tebrates are groups that deserve further study given the

spatial and temporal scarcity of fire effects studies onthese organisms Currently fungi research is highly re-stricted geographically (all studies from ArgentinaLongo et al 2011 2014) while vertebrates showed thelowest number of studies even though this group is fun-damental for controlling cascading effects across trophiclevels (Bruno and Cardinale 2008 Cavallero et al 2013Kurten 2013 Bauer and Hoye 2014 Dirzo et al 2014Peacuterez-Meacutendez et al 2016)Abundance As expected fire negatively affected plant

abundance mainly driven by the impact on woody species(ie shrubs and trees) The negative impact on these

Fig 4 Weighted-mean effect sizes and 95 bias-corrected confidence intervals of fire on abundance from studies published between 1990 and2019 that were used in the review of fire effects across South American ecosystems to determine the role of climate and time since fire Theeffect sizes of overall taxa groups within plant growth forms invertebrates vertebrates climate types and post-fire time are shown Parameterswith confidence intervals that do not overlap the vertical dotted line (Hedgersquos d = 0) are considered to have a significant positive or negativeeffect Sample sizes for each category are shown in parentheses The size of each black dot is proportional to its weight or contribution to theoverall mean calculation Asterisks () denote significant difference (Q-between [Qb] statistics) among categories ( = P lt 00001)

Giorgis et al Fire Ecology (2021) 1711 Page 9 of 20

Fig 5 Weighted-mean effect sizes and 95 bias-corrected confidence intervals of fire on abundance from studies published between 1990 and 2019that were used in the review of fire effects across South American ecosystems to determine the role of climate and time since fire The effect sizes ofearly (-early) and late (-late) post-fire time for abundance of (A) plants invertebrates and vertebrates and (B) grasses shrubs and trees (C) The effectsizes of shrubs trees invertebrates and vertebrates across different climate types (-dry -semiarid -humid-warm) Parameters with confidence intervalsthat do not overlap the vertical dotted line (Hedgersquos d = 0) are considered to have a significant positive or negative effect Sample sizes for eachcategory are shown in parentheses The size of each black dot is proportional to its weight or contribution to the overall mean calculation None ofthe effect sizes of the groups within each moderator variable are significantly different (see Additional file 3 Table 32 for heterogeneity tests)

Giorgis et al Fire Ecology (2021) 1711 Page 10 of 20

growth forms early and late post fire (Fig 5B) is expectedconsidering that the relative recovery of pre-fire abun-dance (including cover and biomass) is slower than inherbaceous species The strongest negative effect observedfor shrubs could be mainly due to the fact that most ofthe studies were carried out in dry-climate ecosystemswhere the combination of water stress and fire signifi-cantly limits shrub recovery (Casillo et al 2012 Pratt et al2014 Jacobsen et al 2016)After fire a combination of different processes may

limit woody species recovery First woody species innon-fire-prone ecosystems may experience high mortal-ity rates (Barlow et al 2003) Second the reduction inaboveground biomass produced by fires exposes individ-uals to direct influence of browsing and trampling bylivestock (Coop et al 2010 Blackhall et al 2015 2017Marcora et al 2018 OrsquoConnor et al 2020 Zeballos et al2020) Third recruitment of woody species seedlingsinto fire-promoted grasslands or herbaceous patchesmay be more restricted due to an increase in competi-tion for light water or nutrients (Scholes and Archer1997 Casillo et al 2012) Finally soil erosion driven byfire increases run-off and seed loss events and reducesnutrient availability therefore constraining biomass pro-duction (Silva et al 2013 Balch et al 2015) The lessernegative effect observed in forbs and grasses may be at-tributed to their relative higher growth rate and shorterlifespan as compared to woody species which result in afaster recovery in burned habitats This vegetativeresponse might be linked to the responses observed insome of the mobile organisms (Fig 4) For example theearly recovery of herbaceous vegetation could promoterapid re-colonization by invertebrates such as insectherbivores and pollinators Additionally our resultsindicated different trends of abundance recovery acrosstaxa groups of invertebrates and vertebrates Similarly aprevious review showed different patterns of ldquooptimumrdquoabundance as time since fire increased (Doherty et al2017) The fast recovery of invertebrate abundance earlypost fire was mainly driven by insect studies (Carboneet al 2019) particularly on Orthoptera (Silveira et al2010 Kral et al 2017) In turn the overall negative effectof fire on the abundance of vertebrates was mainlydriven by bird studies which mostly showed a negativeresponse to fire (Fig 5) Although this pattern has beenobserved in other ecosystems (Fontaine and Kennedy2012 Doherty et al 2017 Carbone et al 2019) the re-sponse of birds to fire can be heterogeneous varyingconsiderably between taxonomic and functional groups(Fontaine and Kennedy 2012) therefore generalizationsmay be difficult to establish The pattern observed herecould be due to the fact that the systematically selectedstudies focused on rather intense fires or on bird speciesthat are more susceptible to fire Our findings of some

Fig 6 Weighted-mean effect sizes and 95 bias-corrected confidenceintervals of fire on plant fitness from studies published between 1990and 2019 that were used in the review of fire effects across SouthAmerican ecosystems to determine the role of climate and time sincefire (A) Effect sizes within growth forms and early and late post-firetime (B) Effect sizes of vegetative and reproductive fitness withingrowth forms (C) Effect sizes of vegetative (-veget) and reproductive(-reprod) fitness within growth forms and early (-early) and late (-late)post-fire time Parameters with confidence intervals that do notoverlap the vertical dotted line (Hedgersquos d = 0) are considered to havea significant positive or negative effect Sample sizes for each categoryare shown in parentheses The size of each black dot is proportional toits weight or contribution to the overall mean calculation Asterisks ()denotes a significant difference (Q-between [Qb] statistics) amongcategories ( = P lt 00001 = P lt 005)

Giorgis et al Fire Ecology (2021) 1711 Page 11 of 20

level of recovery of vertebrate abundance late post firecompared to early post fire (Fig 5) could have beendriven by the re-establishment of some bird populations

over time (Fontaine and Kennedy 2012 Doherty et al2017) Additionally the recovery of populations ofspecialist birds may take longer particularly in forest

Fig 7 Weighted-mean effect sizes and 95 bias-corrected confidence intervals of fire on soil properties from studies published between 1990and 2019 that were used in the review of fire effects across South American ecosystems to determine the role of climate and time since fire (A)Effect sizes of bulk density litter organic matter (OM) total nitrogen (Total N) available nitrogen (N) phosphorous (P) micronutrients pH andsalinity are shown (B) Effect sizes of early (-early) and late (-late) post-fire time of organic matter total nitrogen available nitrogen phosphorousand within climate type (-humid-warm -humid-cold -semiarid) of organic matter and total nitrogen Parameters with confidence intervals that donot overlap the vertical dotted line (Hedgersquos d = 0) are considered to have a significant positive or negative effect Sample sizes for each categoryare shown in parentheses The size of each black dot is proportional to its weight or contribution to the overall mean calculation Asterisk ()denotes a significant difference (Q-between [Qb] statistics) among categories ( = P lt 005)

Giorgis et al Fire Ecology (2021) 1711 Page 12 of 20

ecosystems (eg Barlow and Peres 2004 Fontaine andKennedy 2012 Mestre et al 2013 Albanesi et al 2014)although there are exceptions (eg Fontaine andKennedy 2012 Lee 2018 Morales et al 2020) such asthe group of birds known as cavity excavators (iewoodpeckers [Picidae]) that take advantage of both thesoftness of the wood of recently burned trees (ie snags)for making their cavities (Schepps et al 1999 Winklerand Christie 2002 Bond et al 2012 Lorenz et al 2015)and the higher amount of food (insect and other inverte-brates) found in that substrate right after the fire(Murphy and Lehnhausen 1998 Nappi et al 2003)Plant fitness Despite the negative effect of fire on tree

abundance we found no effect of fire on tree fitnessand a slight positive trend on vegetative fitness early postfire (Fig 6A B) Many tree species in the analyzed eco-systems are highly resilient to individual fires mainlythrough resprouting (Hoffmann and Solbrig 2003 Sou-chie et al 2017 Jaureguiberry et al 2020) therefore ifburned plots do not burn again trees might grow andrecover their biomass (Doherty et al 2017 Miller et al2019 Coop et al 2020) This response does not neces-sarily imply a direct adaptation to fire as the fire-adaptivetraits on which it relies could also have arisen in responseto other factors that have had a long history in some ofthe studied regions such as herbivory and seasonaldrought (Keeley et al 2011) which indirectly provide thevegetation with a high resilience to fire On the otherhand the negative tendency of the reproductive fitness ob-served early post fire is expected since trees may take along time to recover their reproductive size (Chapin IIIet al 2011) In contrast to trees shrubs did show an in-crease in vegetative and reproductive fitness early post firewhich differs significantly from that of late post fire(Fig 6C) The early vegetative response reflects the highresprouting capacity of many shrub species across thestudied regions especially in fire-prone regions (Gurvichet al 2005 Bravo et al 2014 Torres et al 2014 Duriganet al 2020 Jaureguiberry et al 2020) Regarding the repro-ductive response the pattern reported here agrees withprevious evidence suggesting that shrubs might reachminimum reproductive sizes soon after fire then slowdown their growth rate and reproduction as successionprogresses (Hoffmann and Moreira 2002 Hoffmann andSolbrig 2003 Galiacutendez et al 2009) In the case of herb-aceous plants surprisingly we did not find a general effectin fitness but a positive trend in reproductive fitness wasobserved mainly early post fire for grasses (Fig 6B C)The lack of a consistent general response pattern in forbscould be explained by the high heterogeneity of functionalgroups included within this growth form (eg annual per-ennial climber fern) that can have different responses tofire (Keeley et al 1981 Bates et al 2014 Heydari et al2016 Arcamone and Jaureguiberry 2018 Vidaller et al

2019) In the case of grasses the tendency of increasing re-productive fitness early post fire (Fig 6C) could be relatedto the positive firendashcolonization feedback generally de-scribed for this growth form (Bond et al 2003 Bond andKeeley 2005 Pausas and Keeley 2009 Pilon et al 2018)Previous studies had observed an increase in reproductivefitness of grasses after fire (Baruch and Bilbao 1999 Ara-uacutejo et al 2013 Pilon et al 2018 Vidaller et al 2019) Onthe contrary vegetative fitness showed a negative impactearly post fire However we cannot determine if the ob-served pattern (ie positive reproductive fitness versusnegative vegetative fitness) was due to intrinsic differencesbetween types of fitness or to differences in the type of cli-mate since all studies on reproductive fitness were lo-cated in humid-warm climate while vegetative fitnesswas mostly studied in semiarid and arid climatesSoil properties The responses of soil properties to fire in

South American ecosystems support the trend suggestedin previous global reviews (Certini 2005 Pellegrini et al2018) Fire significantly reduces the amount of two im-portant soil properties OM and nitrogen content (Pelle-grini et al 2018) On the one hand OM is directlyassociated with carbon concentration in the soil one ofthe most stable carbon reservoirs worldwide thus large-scale losses of OM due to fire could have important impli-cations for global climate change (Lal 2004) On the otherhand nitrogen is often the most limiting soil nutrient forplant growth and its loss directly affects ecosystems by re-ducing net primary productivity (Vitousek and Howarth1991 Pellegrini et al 2018) In addition our results of ahigher negative effect of fire on OM in semiarid cli-mates compared to humid-warm climate is supportedby Pellegrini et al (2018) who found that across sa-vannas and grasslands biannual fire simulation induceshigh carbon losses under drier climates However ithas been recently proposed that a single fire event maybe related to either an increase a decrease or nochange in soil carbon and nitrogen while frequent firesover time produce a consistent decrease of both ele-ments across ecosystems (Pellegrini et al 2020a b)This hypothesis poses a constraint to our analysessince most of the articles did not report fire history fortheir study sites Consequently the heterogeneous ef-fects of single fires on soil carbon and nitrogen may beconfused with those of frequent fires therefore limitingthe identification of clearer patternsFire produces an increase in bulk density as a result

of soil disaggregation due to the decrease or loss ofOM and to the infiltration of ashes into soil micro-pores (Boyer and Miller 1994 Mataix-Solera et al2011) Soil pH also increased in burned plots as aresult of organic acid denaturation due to soil heatingand the accumulation of potassium and sodiumhydroxides and magnesium and calcium carbonates

Giorgis et al Fire Ecology (2021) 1711 Page 13 of 20

(Knicker 2007) The increase in pH is also driven bythe reduction in OM (Certini 2005) Additionallyimmediate phosphorus enrichment after a fire event isthe result of organic pool conversion of phosphorusinto orthophosphate which declines at later post-firetimes as it is also suggested by Certini (2005)Climatendashfirendashvegetation feedbacks Our review sug-

gests that climatic conditions associated with higherrates of biomass accumulation might determine a fasterecosystem recovery in burned sites across South Amer-ica We found a faster recovery in invertebrate abun-dance and lower losses of soil organic matter underclimates with higher water availability and temperaturesthan in semiarid climates Beyond climatic conditions inhistorically fire-prone ecosystems where fire has pre-sumably been an important evolutionary factor (Bondand Keeley 2005) a greater recovery of the biota is ex-pected compared to non-fire-prone ecosystems Howeverour classification of climate types may not necessarily re-flect the fire history of the corresponding regions there-fore making a clear interpretation of the reported patternsdifficult Although semiarid climate sites could be clearlyconsidered as fire-prone systems (Bond 2005) warm-humid climate sites may also include fire-prone ecosys-tems because some of them have seasonal rainfall pat-terns which is a key characteristic in determining the fireproneness of an ecosystem (Nogueira et al 2017 Romanoand Ursino 2020) In fact following the classification ofOlson et al (2001) 40 of the studies from warm-humidareas corresponded to tropical and subtropical grasslandsavanna and shrublands biome (arguably fire prone)while the other 60 corresponded to tropical and sub-tropical moist broadleaf forest biome (arguably non-fireprone) Most of the former were located in the Cerradoregion where Hoffmann and Moreira (2002) found a fas-ter post-fire recovery of woody species compared toneighboring tropical forest areas Thus they concludedthat the combination of fire-prone vegetation presumablywith a long fire history and warm-humid climate mightfavor a fast ecosystem recovery after fire Therefore theobserved results of greater recovery in warm-humid eco-systems types is likely due to a combination of both cli-mate and fire history Likewise most of the studies fromcold-humid areas have a mediterranean-type climate (egnorthwest Patagonia in Argentina Keeley et al 2012)where rainy winters and hot and dry summers make theseregions fire prone Such regions have been historicallysubjected to periodic fires (both natural and anthropo-genic) and therefore many plant species are capable of re-covering after a fire event (Veblen et al 2003 2008Defosseacute et al 2015) Interestingly we found a significantpost-fire decrease in nitrogen content and soil OM undersemiarid climate but not under cold-humid climateThis response pattern may be due to the inherent

differences in ecosystem properties between climatetypes (eg decomposition and growth rates succes-sional patterns and fire regimen among othersMcLauchlan et al 2020) Fire regime might play animportant role as more frequent surface fire in semi-arid ecosystems could lead to a lower recovery of soilproperties compared to cold-humid ecosystems wherecrown fires predominate and fire frequency is lower thanin semiarid ecosystems therefore buffering fire effects onsoil propertiesThe lack of a standardized classification of fire-

prone and non-fire-prone regions as well as the lim-ited available information on the frequency intensityand severity of fires in the research studies includedin this review are certainly limitations that call forprecaution in the interpretation and extrapolation ofthe results reported here Overcoming such limita-tions remains a challenge in fire ecology (Harris et al2016 Kelly et al 2018) We believe that our studyrepresents a step forward in the synthesis of fire-related patterns at a large scale Our meta-analysis aswell as the discussed patterns provides relevant infor-mation toward understanding the feedback betweenclimate vegetation and fire on the South Americancontinent An ultimately desirable goal in the mediumor long term would be the development of a globalnetwork of fire field experimental surveys across sys-tems with different vegetation types and fire histories

Future perspectivesAcross recently published global fire reviews there isclearly a low representation of studies from SouthAmerica (Prichard et al 2017 Geary et al 2019)Given the relevance of this continent in the globaldynamics of fire it is highly necessary that morestudies on different aspects of fire ecology be carriedout throughout fire-prone regions of the continent(eg seasonal forests and humid and semiarid savannas)Furthermore our review allowed us to identify specific re-search gaps which can be added to the agenda of researchpriorities for future studies in fire ecology (McLauchlanet al 2020)Most of the analyzed studies failed to include fire

characteristics such as frequency intensity and severitywhich prevented us from obtaining a reliable pattern oftheir role as modulators of ecosystem and organismresponses to fire This is an unmet challenge in fireresearch as these fire characteristics may certainly influ-ence the trends reported in this review (Keeley 2009Fontaine and Kennedy 2012 Balch et al 2015 Silveiraet al 2016 Carbone et al 2019)Considering that woody ecosystems might take more

than 40 years to recover after fire (Cavallero et al 2015

Giorgis et al Fire Ecology (2021) 1711 Page 14 of 20

Doherty et al 2017) studies on fire responses that spana short recovery period fail to elucidate what might hap-pen in the meantime between fire occurrence and recov-ery Consequently our review poses an urgent call toincrease the studies assessing changes in fire effects overlonger post-fire times than those usually consideredFurther studies focusing on biodiversity are needed to

increase sensibility to detect changes across taxa groupsclimate types and post-fire timesAdditionally there is a gap of studies on vertebrate

responses to fire and some of the few available ones havemethodological limitations which prevented us from in-cluding them in our analysis (eg lack of replicates inmammals studies Griffiths and Brook 2014)Studies dealing with fire and exotic or invasive species

were quite scarce While invasion ecology has shown in-creasing attention in the last decades (eg Gurevitch et al2011 Pyšek et al 2020) it does not seem to be reflected infire studies in South America (but see Chaneton et al2004 Hoffmann et al 2004 Maziacutea et al 2010 Raffaeleet al 2016 Herrero et al 2016 Marcora et al 2018) Webelieve that this should be a priority topic for future re-search particularly considering the tight link betweenplant invasions and changes in fire regimes (Brooks et al2004 Mandle et al 2011 Harris et al 2016)

Supplementary InformationThe online version contains supplementary material available at httpsdoiorg101186s42408-021-00100-9

Additional file 1 Supplementary material regarding the choice ofstudies published between 1990 and 2019 that were used in the reviewof fire effects across South American ecosystems to determine the role ofclimate and time since fire Figure 11 PRISMA flow diagramrepresenting the flow of information through the decision process (iethe number of studies identified rejected and accepted) for the meta-analysis of fire effects across South American ecosystems to determinethe role of climate and time since fire Studies used were published be-tween 1990 and 2019 List 11 List of reviewed studies published be-tween 1990 and 2019 that were used in the meta-analysis of fire effectsacross South American ecosystems to determine the role of climate andtime since fire

Additional file 2 Distribution across the climatic temperaturendashprecipitation plane of study sites (n = 175) from the 160 articles reviewedfor the meta-analysis of fire effects across South American ecosystems todetermine the role of climate and time since fire Studies used were pub-lished between 1990 and 2019 The four climate types established for thecollected data are indicated by different colors

Additional file 3 Distribution across the climatic temperaturendashprecipitation plane of study sites (n = 175) from the 160 articles reviewedfor the meta-analysis of fire effects across South American ecosystems todetermine the role of climate and time since fire Studies used were pub-lished between 1990 and 2019 The four climate types established for thecollected data are indicated by different colors Table 31 Synopsis ofmodels used in the different hierarchical mixed effects meta-analyses offire effects across South American ecosystems taken from studies pub-lished between 1990 and 2019 to determine the role of climate and timesince fire expressed in R language syntax using the metafor package(Viechtbauer 2010) yi = Hedgesrsquo d vi = variance of Hedgesrsquo d ID = effectidentity Table 32 Summary of the fire effects on biodiversity abun-dance fitness and soil properties showing tests of moderator variables

and heterogeneities from the meta-analyses of fire effects across SouthAmerican ecosystems to determine the role of climate and time sincefire Studies used were published between 1990 and 2019 Significant in-fluence of moderators is indicated in bold when P le 005 and the 95confidence interval (CI) does not include zero Non-significant effects (P gt005) are indicated with ldquonsrdquo Note that Q-total (QT) statistics are reportedfor overall effect and Q-between (Qb) statistics are reported for categor-ical moderator variables

Additional file 4 Supplementary information 41 Publication biasfor the meta-analysis of fire effects across South American ecosystems todetermine the role of climate and time since fire Studies used were pub-lished between 1990 and 2019 Table 41 Sample size (number of obser-vations) and results of publication bias and hierarchical models testingfor each response variable in our meta-analysis of fire effects across SouthAmerican ecosystems to determine the role of climate and time sincefire Studies used were published between 1990 and 2019 Publicationbias was tested using fail-safe number (ie Rosenberg method) Kendallrsquosrank correlation test (z) and ldquotrim and fillrdquo models

AcknowledgementsMAG and HvW were partially supported by the National Scientific andTechnological Research Council (CONICET) fellow program MAG SRZ LC RAAF PAT EK FB DEG PV and PJ are researchers of CONICET Argentina MAGLC PAT EK FB and DEG are professors at the National University of Cordobaand PV is professor at the Universidad Nacional de Cuyo

Authorsrsquo contributionsMAG SRZ LC RA HZ HvW and PJ designed the research and analysesMAG SRZ LC RA and PJ conducted literature search with the contributionof all the authors MAG LC and PJ wrote the article with contributions fromRA AF PT and EK All authors read and approved the final manuscript

FundingThe National Council for Scientific and Technological Research (CONICET) ofArgentina indirectly supported this manuscript through exchange grants forMAG and HvW However no direct funding supported this project

Availability of data and materialsNo unique data were created in the writing of this manuscript so noadditional data are available

Declarations

Ethics approval and consent to participateNot applicable

Consent for publicationConsent for publication not applicable as we use previously published

Competing interestsThe authors declare that they have no competing interests

Author details1Instituto Multidisciplinario de Biologia Vegetal (CONICET-UNC) VelezSarsfield 1611 CC 495 5000 Cordoba Argentina 2Facultad de CienciasExactas Fisicas y Naturales Universidad Nacional de Cordoba Velez Sarsfield299 CC 495 5000 Cordoba Argentina 3Faculty of Sustainability LeuphanaUniversity Luneburg Institute for Ethics and Transdisciplinary SustainabilityResearch Universitatsallee 1 21335 Luneburg Germany 4Faculty ofSustainability Leuphana University Luneburg Institute of EcologyUniversitatsallee 1 21335 Luneburg Germany 5Instituto de Diversidad yEcologia Animal (IDEA) CONICET-UNC and Facultad de Ciencias ExactasFisicas y Naturales Universidad Nacional de Cordoba Velez Sarsfield 299 CP5000 Cordoba Argentina 6Facultad de Ciencias Agrarias UniversidadNacional de Cuyo Brown sin nuacutemero Chacras de Coria 5505 MendozaArgentina 7Facultad de Ciencias Agrarias Universidad Nacional de CuyoM5528AHB Chacras de Coria Mendoza Argentina

Giorgis et al Fire Ecology (2021) 1711 Page 15 of 20

number of 5n+10 (where n is the number of studies in-cluded in the meta-analysis) implying that the resultsfound in our review were robust regardless of the pres-ence of publication bias (Additional file 4 Table 41)

DiscussionThis is the first systematic quantitative review assessingfire effects on multiple taxonomic groups across differ-ent climate types and post-fire times in South Americaone of the worldrsquos most burned continents Our resultsshow non-significant effects of fire on biodiversity asharp negative effect on abundance of woody plant andvertebrate species and an increase in plant fitness ofshrubs In addition we observed clear negative fireeffects in the most important soil properties as waspreviously found at a global scale Our meta-analysisshowed a tendency of stronger negative effect in semi-arid climate than in tropical warm climate indicatingthat higher temperatures and water availability in thelatter climate allow a faster ecosystem recovery after fireAdditionally more significant fire effects were observed

early post fire than late post fire indicating strongereffects immediately after fire and their dilution over time

Effect of fire on the response variablesBiodiversity Biodiversity was not significantly affected byfire across any taxa group climate type or time since fireindicating an overall high resilience at the analyzed spatialand temporal scales This is in line with previous studiesfocused on specific semiarid regions including two well-studied large fire-prone ecosystems in South America (ieCerrado and the Gran Chaco) which had shown post-firerecovery of biodiversity for different taxa groups such asplants birds and insects (eg Kunst et al 2003 2015Kowaljow et al 2019 Durigan et al 2020 Pilon et al2021) However these responses may certainly involvecontext-dependent changes in species composition (Duri-gan et al 2020 McLauchlan et al 2020) which could notbe captured in our quantitative synthesisInterestingly we found a tendency toward higher

invertebrate biodiversity in burned plots compared tocontrols This pattern has been reported in differentareas around the world (eg (Knoechelmann and Morais

Fig 3 Weighted-mean effect sizes and 95 bias-corrected confidence intervals of fire on from studies published between 1990 and 2019 thatwere used in the review of fire effects across South American ecosystems to determine the role of climate and time since fire The effect sizes ofoverall biodiversity taxa groups climate post-fire time and the response of plants in early (plants-early) and late (plants-late) post fire are shownParameters with confidence intervals that do not overlap the vertical dotted line (Hedgersquos d = 0) are considered to have a significant positive ornegative effect Sample sizes for each category are shown in parentheses The size of each black dot is proportional to its weight or contributionto the overall mean calculation

Giorgis et al Fire Ecology (2021) 1711 Page 8 of 20

2008 Uehara-Prado et al 2010 Carbone et al 2019Lazarina et al 2019) and could be explained by the highmobility of many invertebrates (eg arthropods) whichcan take advantage of the usually high availability ofplant biomass in early post-fire succession (Mirandaet al 2002 Hoffmann and Andersen 2003) Contrary toinvertebrates diversity of fungi (ie mycorrhizas) tendedto decrease (Fig 3) probably as a response to the glo-bally general negative effect of fire on soil microbialcommunities (Pressler et al 2019) and soil quality (Cer-tini 2005 Pellegrini et al 2018) However fungi and ver-tebrates are groups that deserve further study given the

spatial and temporal scarcity of fire effects studies onthese organisms Currently fungi research is highly re-stricted geographically (all studies from ArgentinaLongo et al 2011 2014) while vertebrates showed thelowest number of studies even though this group is fun-damental for controlling cascading effects across trophiclevels (Bruno and Cardinale 2008 Cavallero et al 2013Kurten 2013 Bauer and Hoye 2014 Dirzo et al 2014Peacuterez-Meacutendez et al 2016)Abundance As expected fire negatively affected plant

abundance mainly driven by the impact on woody species(ie shrubs and trees) The negative impact on these

Fig 4 Weighted-mean effect sizes and 95 bias-corrected confidence intervals of fire on abundance from studies published between 1990 and2019 that were used in the review of fire effects across South American ecosystems to determine the role of climate and time since fire Theeffect sizes of overall taxa groups within plant growth forms invertebrates vertebrates climate types and post-fire time are shown Parameterswith confidence intervals that do not overlap the vertical dotted line (Hedgersquos d = 0) are considered to have a significant positive or negativeeffect Sample sizes for each category are shown in parentheses The size of each black dot is proportional to its weight or contribution to theoverall mean calculation Asterisks () denote significant difference (Q-between [Qb] statistics) among categories ( = P lt 00001)

Giorgis et al Fire Ecology (2021) 1711 Page 9 of 20

Fig 5 Weighted-mean effect sizes and 95 bias-corrected confidence intervals of fire on abundance from studies published between 1990 and 2019that were used in the review of fire effects across South American ecosystems to determine the role of climate and time since fire The effect sizes ofearly (-early) and late (-late) post-fire time for abundance of (A) plants invertebrates and vertebrates and (B) grasses shrubs and trees (C) The effectsizes of shrubs trees invertebrates and vertebrates across different climate types (-dry -semiarid -humid-warm) Parameters with confidence intervalsthat do not overlap the vertical dotted line (Hedgersquos d = 0) are considered to have a significant positive or negative effect Sample sizes for eachcategory are shown in parentheses The size of each black dot is proportional to its weight or contribution to the overall mean calculation None ofthe effect sizes of the groups within each moderator variable are significantly different (see Additional file 3 Table 32 for heterogeneity tests)

Giorgis et al Fire Ecology (2021) 1711 Page 10 of 20

growth forms early and late post fire (Fig 5B) is expectedconsidering that the relative recovery of pre-fire abun-dance (including cover and biomass) is slower than inherbaceous species The strongest negative effect observedfor shrubs could be mainly due to the fact that most ofthe studies were carried out in dry-climate ecosystemswhere the combination of water stress and fire signifi-cantly limits shrub recovery (Casillo et al 2012 Pratt et al2014 Jacobsen et al 2016)After fire a combination of different processes may

limit woody species recovery First woody species innon-fire-prone ecosystems may experience high mortal-ity rates (Barlow et al 2003) Second the reduction inaboveground biomass produced by fires exposes individ-uals to direct influence of browsing and trampling bylivestock (Coop et al 2010 Blackhall et al 2015 2017Marcora et al 2018 OrsquoConnor et al 2020 Zeballos et al2020) Third recruitment of woody species seedlingsinto fire-promoted grasslands or herbaceous patchesmay be more restricted due to an increase in competi-tion for light water or nutrients (Scholes and Archer1997 Casillo et al 2012) Finally soil erosion driven byfire increases run-off and seed loss events and reducesnutrient availability therefore constraining biomass pro-duction (Silva et al 2013 Balch et al 2015) The lessernegative effect observed in forbs and grasses may be at-tributed to their relative higher growth rate and shorterlifespan as compared to woody species which result in afaster recovery in burned habitats This vegetativeresponse might be linked to the responses observed insome of the mobile organisms (Fig 4) For example theearly recovery of herbaceous vegetation could promoterapid re-colonization by invertebrates such as insectherbivores and pollinators Additionally our resultsindicated different trends of abundance recovery acrosstaxa groups of invertebrates and vertebrates Similarly aprevious review showed different patterns of ldquooptimumrdquoabundance as time since fire increased (Doherty et al2017) The fast recovery of invertebrate abundance earlypost fire was mainly driven by insect studies (Carboneet al 2019) particularly on Orthoptera (Silveira et al2010 Kral et al 2017) In turn the overall negative effectof fire on the abundance of vertebrates was mainlydriven by bird studies which mostly showed a negativeresponse to fire (Fig 5) Although this pattern has beenobserved in other ecosystems (Fontaine and Kennedy2012 Doherty et al 2017 Carbone et al 2019) the re-sponse of birds to fire can be heterogeneous varyingconsiderably between taxonomic and functional groups(Fontaine and Kennedy 2012) therefore generalizationsmay be difficult to establish The pattern observed herecould be due to the fact that the systematically selectedstudies focused on rather intense fires or on bird speciesthat are more susceptible to fire Our findings of some

Fig 6 Weighted-mean effect sizes and 95 bias-corrected confidenceintervals of fire on plant fitness from studies published between 1990and 2019 that were used in the review of fire effects across SouthAmerican ecosystems to determine the role of climate and time sincefire (A) Effect sizes within growth forms and early and late post-firetime (B) Effect sizes of vegetative and reproductive fitness withingrowth forms (C) Effect sizes of vegetative (-veget) and reproductive(-reprod) fitness within growth forms and early (-early) and late (-late)post-fire time Parameters with confidence intervals that do notoverlap the vertical dotted line (Hedgersquos d = 0) are considered to havea significant positive or negative effect Sample sizes for each categoryare shown in parentheses The size of each black dot is proportional toits weight or contribution to the overall mean calculation Asterisks ()denotes a significant difference (Q-between [Qb] statistics) amongcategories ( = P lt 00001 = P lt 005)

Giorgis et al Fire Ecology (2021) 1711 Page 11 of 20

level of recovery of vertebrate abundance late post firecompared to early post fire (Fig 5) could have beendriven by the re-establishment of some bird populations

over time (Fontaine and Kennedy 2012 Doherty et al2017) Additionally the recovery of populations ofspecialist birds may take longer particularly in forest

Fig 7 Weighted-mean effect sizes and 95 bias-corrected confidence intervals of fire on soil properties from studies published between 1990and 2019 that were used in the review of fire effects across South American ecosystems to determine the role of climate and time since fire (A)Effect sizes of bulk density litter organic matter (OM) total nitrogen (Total N) available nitrogen (N) phosphorous (P) micronutrients pH andsalinity are shown (B) Effect sizes of early (-early) and late (-late) post-fire time of organic matter total nitrogen available nitrogen phosphorousand within climate type (-humid-warm -humid-cold -semiarid) of organic matter and total nitrogen Parameters with confidence intervals that donot overlap the vertical dotted line (Hedgersquos d = 0) are considered to have a significant positive or negative effect Sample sizes for each categoryare shown in parentheses The size of each black dot is proportional to its weight or contribution to the overall mean calculation Asterisk ()denotes a significant difference (Q-between [Qb] statistics) among categories ( = P lt 005)

Giorgis et al Fire Ecology (2021) 1711 Page 12 of 20

ecosystems (eg Barlow and Peres 2004 Fontaine andKennedy 2012 Mestre et al 2013 Albanesi et al 2014)although there are exceptions (eg Fontaine andKennedy 2012 Lee 2018 Morales et al 2020) such asthe group of birds known as cavity excavators (iewoodpeckers [Picidae]) that take advantage of both thesoftness of the wood of recently burned trees (ie snags)for making their cavities (Schepps et al 1999 Winklerand Christie 2002 Bond et al 2012 Lorenz et al 2015)and the higher amount of food (insect and other inverte-brates) found in that substrate right after the fire(Murphy and Lehnhausen 1998 Nappi et al 2003)Plant fitness Despite the negative effect of fire on tree

abundance we found no effect of fire on tree fitnessand a slight positive trend on vegetative fitness early postfire (Fig 6A B) Many tree species in the analyzed eco-systems are highly resilient to individual fires mainlythrough resprouting (Hoffmann and Solbrig 2003 Sou-chie et al 2017 Jaureguiberry et al 2020) therefore ifburned plots do not burn again trees might grow andrecover their biomass (Doherty et al 2017 Miller et al2019 Coop et al 2020) This response does not neces-sarily imply a direct adaptation to fire as the fire-adaptivetraits on which it relies could also have arisen in responseto other factors that have had a long history in some ofthe studied regions such as herbivory and seasonaldrought (Keeley et al 2011) which indirectly provide thevegetation with a high resilience to fire On the otherhand the negative tendency of the reproductive fitness ob-served early post fire is expected since trees may take along time to recover their reproductive size (Chapin IIIet al 2011) In contrast to trees shrubs did show an in-crease in vegetative and reproductive fitness early post firewhich differs significantly from that of late post fire(Fig 6C) The early vegetative response reflects the highresprouting capacity of many shrub species across thestudied regions especially in fire-prone regions (Gurvichet al 2005 Bravo et al 2014 Torres et al 2014 Duriganet al 2020 Jaureguiberry et al 2020) Regarding the repro-ductive response the pattern reported here agrees withprevious evidence suggesting that shrubs might reachminimum reproductive sizes soon after fire then slowdown their growth rate and reproduction as successionprogresses (Hoffmann and Moreira 2002 Hoffmann andSolbrig 2003 Galiacutendez et al 2009) In the case of herb-aceous plants surprisingly we did not find a general effectin fitness but a positive trend in reproductive fitness wasobserved mainly early post fire for grasses (Fig 6B C)The lack of a consistent general response pattern in forbscould be explained by the high heterogeneity of functionalgroups included within this growth form (eg annual per-ennial climber fern) that can have different responses tofire (Keeley et al 1981 Bates et al 2014 Heydari et al2016 Arcamone and Jaureguiberry 2018 Vidaller et al

2019) In the case of grasses the tendency of increasing re-productive fitness early post fire (Fig 6C) could be relatedto the positive firendashcolonization feedback generally de-scribed for this growth form (Bond et al 2003 Bond andKeeley 2005 Pausas and Keeley 2009 Pilon et al 2018)Previous studies had observed an increase in reproductivefitness of grasses after fire (Baruch and Bilbao 1999 Ara-uacutejo et al 2013 Pilon et al 2018 Vidaller et al 2019) Onthe contrary vegetative fitness showed a negative impactearly post fire However we cannot determine if the ob-served pattern (ie positive reproductive fitness versusnegative vegetative fitness) was due to intrinsic differencesbetween types of fitness or to differences in the type of cli-mate since all studies on reproductive fitness were lo-cated in humid-warm climate while vegetative fitnesswas mostly studied in semiarid and arid climatesSoil properties The responses of soil properties to fire in

South American ecosystems support the trend suggestedin previous global reviews (Certini 2005 Pellegrini et al2018) Fire significantly reduces the amount of two im-portant soil properties OM and nitrogen content (Pelle-grini et al 2018) On the one hand OM is directlyassociated with carbon concentration in the soil one ofthe most stable carbon reservoirs worldwide thus large-scale losses of OM due to fire could have important impli-cations for global climate change (Lal 2004) On the otherhand nitrogen is often the most limiting soil nutrient forplant growth and its loss directly affects ecosystems by re-ducing net primary productivity (Vitousek and Howarth1991 Pellegrini et al 2018) In addition our results of ahigher negative effect of fire on OM in semiarid cli-mates compared to humid-warm climate is supportedby Pellegrini et al (2018) who found that across sa-vannas and grasslands biannual fire simulation induceshigh carbon losses under drier climates However ithas been recently proposed that a single fire event maybe related to either an increase a decrease or nochange in soil carbon and nitrogen while frequent firesover time produce a consistent decrease of both ele-ments across ecosystems (Pellegrini et al 2020a b)This hypothesis poses a constraint to our analysessince most of the articles did not report fire history fortheir study sites Consequently the heterogeneous ef-fects of single fires on soil carbon and nitrogen may beconfused with those of frequent fires therefore limitingthe identification of clearer patternsFire produces an increase in bulk density as a result

of soil disaggregation due to the decrease or loss ofOM and to the infiltration of ashes into soil micro-pores (Boyer and Miller 1994 Mataix-Solera et al2011) Soil pH also increased in burned plots as aresult of organic acid denaturation due to soil heatingand the accumulation of potassium and sodiumhydroxides and magnesium and calcium carbonates

Giorgis et al Fire Ecology (2021) 1711 Page 13 of 20

(Knicker 2007) The increase in pH is also driven bythe reduction in OM (Certini 2005) Additionallyimmediate phosphorus enrichment after a fire event isthe result of organic pool conversion of phosphorusinto orthophosphate which declines at later post-firetimes as it is also suggested by Certini (2005)Climatendashfirendashvegetation feedbacks Our review sug-

gests that climatic conditions associated with higherrates of biomass accumulation might determine a fasterecosystem recovery in burned sites across South Amer-ica We found a faster recovery in invertebrate abun-dance and lower losses of soil organic matter underclimates with higher water availability and temperaturesthan in semiarid climates Beyond climatic conditions inhistorically fire-prone ecosystems where fire has pre-sumably been an important evolutionary factor (Bondand Keeley 2005) a greater recovery of the biota is ex-pected compared to non-fire-prone ecosystems Howeverour classification of climate types may not necessarily re-flect the fire history of the corresponding regions there-fore making a clear interpretation of the reported patternsdifficult Although semiarid climate sites could be clearlyconsidered as fire-prone systems (Bond 2005) warm-humid climate sites may also include fire-prone ecosys-tems because some of them have seasonal rainfall pat-terns which is a key characteristic in determining the fireproneness of an ecosystem (Nogueira et al 2017 Romanoand Ursino 2020) In fact following the classification ofOlson et al (2001) 40 of the studies from warm-humidareas corresponded to tropical and subtropical grasslandsavanna and shrublands biome (arguably fire prone)while the other 60 corresponded to tropical and sub-tropical moist broadleaf forest biome (arguably non-fireprone) Most of the former were located in the Cerradoregion where Hoffmann and Moreira (2002) found a fas-ter post-fire recovery of woody species compared toneighboring tropical forest areas Thus they concludedthat the combination of fire-prone vegetation presumablywith a long fire history and warm-humid climate mightfavor a fast ecosystem recovery after fire Therefore theobserved results of greater recovery in warm-humid eco-systems types is likely due to a combination of both cli-mate and fire history Likewise most of the studies fromcold-humid areas have a mediterranean-type climate (egnorthwest Patagonia in Argentina Keeley et al 2012)where rainy winters and hot and dry summers make theseregions fire prone Such regions have been historicallysubjected to periodic fires (both natural and anthropo-genic) and therefore many plant species are capable of re-covering after a fire event (Veblen et al 2003 2008Defosseacute et al 2015) Interestingly we found a significantpost-fire decrease in nitrogen content and soil OM undersemiarid climate but not under cold-humid climateThis response pattern may be due to the inherent

differences in ecosystem properties between climatetypes (eg decomposition and growth rates succes-sional patterns and fire regimen among othersMcLauchlan et al 2020) Fire regime might play animportant role as more frequent surface fire in semi-arid ecosystems could lead to a lower recovery of soilproperties compared to cold-humid ecosystems wherecrown fires predominate and fire frequency is lower thanin semiarid ecosystems therefore buffering fire effects onsoil propertiesThe lack of a standardized classification of fire-

prone and non-fire-prone regions as well as the lim-ited available information on the frequency intensityand severity of fires in the research studies includedin this review are certainly limitations that call forprecaution in the interpretation and extrapolation ofthe results reported here Overcoming such limita-tions remains a challenge in fire ecology (Harris et al2016 Kelly et al 2018) We believe that our studyrepresents a step forward in the synthesis of fire-related patterns at a large scale Our meta-analysis aswell as the discussed patterns provides relevant infor-mation toward understanding the feedback betweenclimate vegetation and fire on the South Americancontinent An ultimately desirable goal in the mediumor long term would be the development of a globalnetwork of fire field experimental surveys across sys-tems with different vegetation types and fire histories

Future perspectivesAcross recently published global fire reviews there isclearly a low representation of studies from SouthAmerica (Prichard et al 2017 Geary et al 2019)Given the relevance of this continent in the globaldynamics of fire it is highly necessary that morestudies on different aspects of fire ecology be carriedout throughout fire-prone regions of the continent(eg seasonal forests and humid and semiarid savannas)Furthermore our review allowed us to identify specific re-search gaps which can be added to the agenda of researchpriorities for future studies in fire ecology (McLauchlanet al 2020)Most of the analyzed studies failed to include fire

characteristics such as frequency intensity and severitywhich prevented us from obtaining a reliable pattern oftheir role as modulators of ecosystem and organismresponses to fire This is an unmet challenge in fireresearch as these fire characteristics may certainly influ-ence the trends reported in this review (Keeley 2009Fontaine and Kennedy 2012 Balch et al 2015 Silveiraet al 2016 Carbone et al 2019)Considering that woody ecosystems might take more

than 40 years to recover after fire (Cavallero et al 2015

Giorgis et al Fire Ecology (2021) 1711 Page 14 of 20

Doherty et al 2017) studies on fire responses that spana short recovery period fail to elucidate what might hap-pen in the meantime between fire occurrence and recov-ery Consequently our review poses an urgent call toincrease the studies assessing changes in fire effects overlonger post-fire times than those usually consideredFurther studies focusing on biodiversity are needed to

increase sensibility to detect changes across taxa groupsclimate types and post-fire timesAdditionally there is a gap of studies on vertebrate

responses to fire and some of the few available ones havemethodological limitations which prevented us from in-cluding them in our analysis (eg lack of replicates inmammals studies Griffiths and Brook 2014)Studies dealing with fire and exotic or invasive species

were quite scarce While invasion ecology has shown in-creasing attention in the last decades (eg Gurevitch et al2011 Pyšek et al 2020) it does not seem to be reflected infire studies in South America (but see Chaneton et al2004 Hoffmann et al 2004 Maziacutea et al 2010 Raffaeleet al 2016 Herrero et al 2016 Marcora et al 2018) Webelieve that this should be a priority topic for future re-search particularly considering the tight link betweenplant invasions and changes in fire regimes (Brooks et al2004 Mandle et al 2011 Harris et al 2016)

Supplementary InformationThe online version contains supplementary material available at httpsdoiorg101186s42408-021-00100-9

Additional file 1 Supplementary material regarding the choice ofstudies published between 1990 and 2019 that were used in the reviewof fire effects across South American ecosystems to determine the role ofclimate and time since fire Figure 11 PRISMA flow diagramrepresenting the flow of information through the decision process (iethe number of studies identified rejected and accepted) for the meta-analysis of fire effects across South American ecosystems to determinethe role of climate and time since fire Studies used were published be-tween 1990 and 2019 List 11 List of reviewed studies published be-tween 1990 and 2019 that were used in the meta-analysis of fire effectsacross South American ecosystems to determine the role of climate andtime since fire

Additional file 2 Distribution across the climatic temperaturendashprecipitation plane of study sites (n = 175) from the 160 articles reviewedfor the meta-analysis of fire effects across South American ecosystems todetermine the role of climate and time since fire Studies used were pub-lished between 1990 and 2019 The four climate types established for thecollected data are indicated by different colors

Additional file 3 Distribution across the climatic temperaturendashprecipitation plane of study sites (n = 175) from the 160 articles reviewedfor the meta-analysis of fire effects across South American ecosystems todetermine the role of climate and time since fire Studies used were pub-lished between 1990 and 2019 The four climate types established for thecollected data are indicated by different colors Table 31 Synopsis ofmodels used in the different hierarchical mixed effects meta-analyses offire effects across South American ecosystems taken from studies pub-lished between 1990 and 2019 to determine the role of climate and timesince fire expressed in R language syntax using the metafor package(Viechtbauer 2010) yi = Hedgesrsquo d vi = variance of Hedgesrsquo d ID = effectidentity Table 32 Summary of the fire effects on biodiversity abun-dance fitness and soil properties showing tests of moderator variables

and heterogeneities from the meta-analyses of fire effects across SouthAmerican ecosystems to determine the role of climate and time sincefire Studies used were published between 1990 and 2019 Significant in-fluence of moderators is indicated in bold when P le 005 and the 95confidence interval (CI) does not include zero Non-significant effects (P gt005) are indicated with ldquonsrdquo Note that Q-total (QT) statistics are reportedfor overall effect and Q-between (Qb) statistics are reported for categor-ical moderator variables

Additional file 4 Supplementary information 41 Publication biasfor the meta-analysis of fire effects across South American ecosystems todetermine the role of climate and time since fire Studies used were pub-lished between 1990 and 2019 Table 41 Sample size (number of obser-vations) and results of publication bias and hierarchical models testingfor each response variable in our meta-analysis of fire effects across SouthAmerican ecosystems to determine the role of climate and time sincefire Studies used were published between 1990 and 2019 Publicationbias was tested using fail-safe number (ie Rosenberg method) Kendallrsquosrank correlation test (z) and ldquotrim and fillrdquo models

AcknowledgementsMAG and HvW were partially supported by the National Scientific andTechnological Research Council (CONICET) fellow program MAG SRZ LC RAAF PAT EK FB DEG PV and PJ are researchers of CONICET Argentina MAGLC PAT EK FB and DEG are professors at the National University of Cordobaand PV is professor at the Universidad Nacional de Cuyo

Authorsrsquo contributionsMAG SRZ LC RA HZ HvW and PJ designed the research and analysesMAG SRZ LC RA and PJ conducted literature search with the contributionof all the authors MAG LC and PJ wrote the article with contributions fromRA AF PT and EK All authors read and approved the final manuscript

FundingThe National Council for Scientific and Technological Research (CONICET) ofArgentina indirectly supported this manuscript through exchange grants forMAG and HvW However no direct funding supported this project

Availability of data and materialsNo unique data were created in the writing of this manuscript so noadditional data are available

Declarations

Ethics approval and consent to participateNot applicable

Consent for publicationConsent for publication not applicable as we use previously published

Competing interestsThe authors declare that they have no competing interests

Author details1Instituto Multidisciplinario de Biologia Vegetal (CONICET-UNC) VelezSarsfield 1611 CC 495 5000 Cordoba Argentina 2Facultad de CienciasExactas Fisicas y Naturales Universidad Nacional de Cordoba Velez Sarsfield299 CC 495 5000 Cordoba Argentina 3Faculty of Sustainability LeuphanaUniversity Luneburg Institute for Ethics and Transdisciplinary SustainabilityResearch Universitatsallee 1 21335 Luneburg Germany 4Faculty ofSustainability Leuphana University Luneburg Institute of EcologyUniversitatsallee 1 21335 Luneburg Germany 5Instituto de Diversidad yEcologia Animal (IDEA) CONICET-UNC and Facultad de Ciencias ExactasFisicas y Naturales Universidad Nacional de Cordoba Velez Sarsfield 299 CP5000 Cordoba Argentina 6Facultad de Ciencias Agrarias UniversidadNacional de Cuyo Brown sin nuacutemero Chacras de Coria 5505 MendozaArgentina 7Facultad de Ciencias Agrarias Universidad Nacional de CuyoM5528AHB Chacras de Coria Mendoza Argentina

Giorgis et al Fire Ecology (2021) 1711 Page 15 of 20

2008 Uehara-Prado et al 2010 Carbone et al 2019Lazarina et al 2019) and could be explained by the highmobility of many invertebrates (eg arthropods) whichcan take advantage of the usually high availability ofplant biomass in early post-fire succession (Mirandaet al 2002 Hoffmann and Andersen 2003) Contrary toinvertebrates diversity of fungi (ie mycorrhizas) tendedto decrease (Fig 3) probably as a response to the glo-bally general negative effect of fire on soil microbialcommunities (Pressler et al 2019) and soil quality (Cer-tini 2005 Pellegrini et al 2018) However fungi and ver-tebrates are groups that deserve further study given the

spatial and temporal scarcity of fire effects studies onthese organisms Currently fungi research is highly re-stricted geographically (all studies from ArgentinaLongo et al 2011 2014) while vertebrates showed thelowest number of studies even though this group is fun-damental for controlling cascading effects across trophiclevels (Bruno and Cardinale 2008 Cavallero et al 2013Kurten 2013 Bauer and Hoye 2014 Dirzo et al 2014Peacuterez-Meacutendez et al 2016)Abundance As expected fire negatively affected plant

abundance mainly driven by the impact on woody species(ie shrubs and trees) The negative impact on these

Fig 4 Weighted-mean effect sizes and 95 bias-corrected confidence intervals of fire on abundance from studies published between 1990 and2019 that were used in the review of fire effects across South American ecosystems to determine the role of climate and time since fire Theeffect sizes of overall taxa groups within plant growth forms invertebrates vertebrates climate types and post-fire time are shown Parameterswith confidence intervals that do not overlap the vertical dotted line (Hedgersquos d = 0) are considered to have a significant positive or negativeeffect Sample sizes for each category are shown in parentheses The size of each black dot is proportional to its weight or contribution to theoverall mean calculation Asterisks () denote significant difference (Q-between [Qb] statistics) among categories ( = P lt 00001)

Giorgis et al Fire Ecology (2021) 1711 Page 9 of 20

Fig 5 Weighted-mean effect sizes and 95 bias-corrected confidence intervals of fire on abundance from studies published between 1990 and 2019that were used in the review of fire effects across South American ecosystems to determine the role of climate and time since fire The effect sizes ofearly (-early) and late (-late) post-fire time for abundance of (A) plants invertebrates and vertebrates and (B) grasses shrubs and trees (C) The effectsizes of shrubs trees invertebrates and vertebrates across different climate types (-dry -semiarid -humid-warm) Parameters with confidence intervalsthat do not overlap the vertical dotted line (Hedgersquos d = 0) are considered to have a significant positive or negative effect Sample sizes for eachcategory are shown in parentheses The size of each black dot is proportional to its weight or contribution to the overall mean calculation None ofthe effect sizes of the groups within each moderator variable are significantly different (see Additional file 3 Table 32 for heterogeneity tests)

Giorgis et al Fire Ecology (2021) 1711 Page 10 of 20

growth forms early and late post fire (Fig 5B) is expectedconsidering that the relative recovery of pre-fire abun-dance (including cover and biomass) is slower than inherbaceous species The strongest negative effect observedfor shrubs could be mainly due to the fact that most ofthe studies were carried out in dry-climate ecosystemswhere the combination of water stress and fire signifi-cantly limits shrub recovery (Casillo et al 2012 Pratt et al2014 Jacobsen et al 2016)After fire a combination of different processes may

limit woody species recovery First woody species innon-fire-prone ecosystems may experience high mortal-ity rates (Barlow et al 2003) Second the reduction inaboveground biomass produced by fires exposes individ-uals to direct influence of browsing and trampling bylivestock (Coop et al 2010 Blackhall et al 2015 2017Marcora et al 2018 OrsquoConnor et al 2020 Zeballos et al2020) Third recruitment of woody species seedlingsinto fire-promoted grasslands or herbaceous patchesmay be more restricted due to an increase in competi-tion for light water or nutrients (Scholes and Archer1997 Casillo et al 2012) Finally soil erosion driven byfire increases run-off and seed loss events and reducesnutrient availability therefore constraining biomass pro-duction (Silva et al 2013 Balch et al 2015) The lessernegative effect observed in forbs and grasses may be at-tributed to their relative higher growth rate and shorterlifespan as compared to woody species which result in afaster recovery in burned habitats This vegetativeresponse might be linked to the responses observed insome of the mobile organisms (Fig 4) For example theearly recovery of herbaceous vegetation could promoterapid re-colonization by invertebrates such as insectherbivores and pollinators Additionally our resultsindicated different trends of abundance recovery acrosstaxa groups of invertebrates and vertebrates Similarly aprevious review showed different patterns of ldquooptimumrdquoabundance as time since fire increased (Doherty et al2017) The fast recovery of invertebrate abundance earlypost fire was mainly driven by insect studies (Carboneet al 2019) particularly on Orthoptera (Silveira et al2010 Kral et al 2017) In turn the overall negative effectof fire on the abundance of vertebrates was mainlydriven by bird studies which mostly showed a negativeresponse to fire (Fig 5) Although this pattern has beenobserved in other ecosystems (Fontaine and Kennedy2012 Doherty et al 2017 Carbone et al 2019) the re-sponse of birds to fire can be heterogeneous varyingconsiderably between taxonomic and functional groups(Fontaine and Kennedy 2012) therefore generalizationsmay be difficult to establish The pattern observed herecould be due to the fact that the systematically selectedstudies focused on rather intense fires or on bird speciesthat are more susceptible to fire Our findings of some

Fig 6 Weighted-mean effect sizes and 95 bias-corrected confidenceintervals of fire on plant fitness from studies published between 1990and 2019 that were used in the review of fire effects across SouthAmerican ecosystems to determine the role of climate and time sincefire (A) Effect sizes within growth forms and early and late post-firetime (B) Effect sizes of vegetative and reproductive fitness withingrowth forms (C) Effect sizes of vegetative (-veget) and reproductive(-reprod) fitness within growth forms and early (-early) and late (-late)post-fire time Parameters with confidence intervals that do notoverlap the vertical dotted line (Hedgersquos d = 0) are considered to havea significant positive or negative effect Sample sizes for each categoryare shown in parentheses The size of each black dot is proportional toits weight or contribution to the overall mean calculation Asterisks ()denotes a significant difference (Q-between [Qb] statistics) amongcategories ( = P lt 00001 = P lt 005)

Giorgis et al Fire Ecology (2021) 1711 Page 11 of 20

level of recovery of vertebrate abundance late post firecompared to early post fire (Fig 5) could have beendriven by the re-establishment of some bird populations

over time (Fontaine and Kennedy 2012 Doherty et al2017) Additionally the recovery of populations ofspecialist birds may take longer particularly in forest

Fig 7 Weighted-mean effect sizes and 95 bias-corrected confidence intervals of fire on soil properties from studies published between 1990and 2019 that were used in the review of fire effects across South American ecosystems to determine the role of climate and time since fire (A)Effect sizes of bulk density litter organic matter (OM) total nitrogen (Total N) available nitrogen (N) phosphorous (P) micronutrients pH andsalinity are shown (B) Effect sizes of early (-early) and late (-late) post-fire time of organic matter total nitrogen available nitrogen phosphorousand within climate type (-humid-warm -humid-cold -semiarid) of organic matter and total nitrogen Parameters with confidence intervals that donot overlap the vertical dotted line (Hedgersquos d = 0) are considered to have a significant positive or negative effect Sample sizes for each categoryare shown in parentheses The size of each black dot is proportional to its weight or contribution to the overall mean calculation Asterisk ()denotes a significant difference (Q-between [Qb] statistics) among categories ( = P lt 005)

Giorgis et al Fire Ecology (2021) 1711 Page 12 of 20

ecosystems (eg Barlow and Peres 2004 Fontaine andKennedy 2012 Mestre et al 2013 Albanesi et al 2014)although there are exceptions (eg Fontaine andKennedy 2012 Lee 2018 Morales et al 2020) such asthe group of birds known as cavity excavators (iewoodpeckers [Picidae]) that take advantage of both thesoftness of the wood of recently burned trees (ie snags)for making their cavities (Schepps et al 1999 Winklerand Christie 2002 Bond et al 2012 Lorenz et al 2015)and the higher amount of food (insect and other inverte-brates) found in that substrate right after the fire(Murphy and Lehnhausen 1998 Nappi et al 2003)Plant fitness Despite the negative effect of fire on tree

abundance we found no effect of fire on tree fitnessand a slight positive trend on vegetative fitness early postfire (Fig 6A B) Many tree species in the analyzed eco-systems are highly resilient to individual fires mainlythrough resprouting (Hoffmann and Solbrig 2003 Sou-chie et al 2017 Jaureguiberry et al 2020) therefore ifburned plots do not burn again trees might grow andrecover their biomass (Doherty et al 2017 Miller et al2019 Coop et al 2020) This response does not neces-sarily imply a direct adaptation to fire as the fire-adaptivetraits on which it relies could also have arisen in responseto other factors that have had a long history in some ofthe studied regions such as herbivory and seasonaldrought (Keeley et al 2011) which indirectly provide thevegetation with a high resilience to fire On the otherhand the negative tendency of the reproductive fitness ob-served early post fire is expected since trees may take along time to recover their reproductive size (Chapin IIIet al 2011) In contrast to trees shrubs did show an in-crease in vegetative and reproductive fitness early post firewhich differs significantly from that of late post fire(Fig 6C) The early vegetative response reflects the highresprouting capacity of many shrub species across thestudied regions especially in fire-prone regions (Gurvichet al 2005 Bravo et al 2014 Torres et al 2014 Duriganet al 2020 Jaureguiberry et al 2020) Regarding the repro-ductive response the pattern reported here agrees withprevious evidence suggesting that shrubs might reachminimum reproductive sizes soon after fire then slowdown their growth rate and reproduction as successionprogresses (Hoffmann and Moreira 2002 Hoffmann andSolbrig 2003 Galiacutendez et al 2009) In the case of herb-aceous plants surprisingly we did not find a general effectin fitness but a positive trend in reproductive fitness wasobserved mainly early post fire for grasses (Fig 6B C)The lack of a consistent general response pattern in forbscould be explained by the high heterogeneity of functionalgroups included within this growth form (eg annual per-ennial climber fern) that can have different responses tofire (Keeley et al 1981 Bates et al 2014 Heydari et al2016 Arcamone and Jaureguiberry 2018 Vidaller et al

2019) In the case of grasses the tendency of increasing re-productive fitness early post fire (Fig 6C) could be relatedto the positive firendashcolonization feedback generally de-scribed for this growth form (Bond et al 2003 Bond andKeeley 2005 Pausas and Keeley 2009 Pilon et al 2018)Previous studies had observed an increase in reproductivefitness of grasses after fire (Baruch and Bilbao 1999 Ara-uacutejo et al 2013 Pilon et al 2018 Vidaller et al 2019) Onthe contrary vegetative fitness showed a negative impactearly post fire However we cannot determine if the ob-served pattern (ie positive reproductive fitness versusnegative vegetative fitness) was due to intrinsic differencesbetween types of fitness or to differences in the type of cli-mate since all studies on reproductive fitness were lo-cated in humid-warm climate while vegetative fitnesswas mostly studied in semiarid and arid climatesSoil properties The responses of soil properties to fire in

South American ecosystems support the trend suggestedin previous global reviews (Certini 2005 Pellegrini et al2018) Fire significantly reduces the amount of two im-portant soil properties OM and nitrogen content (Pelle-grini et al 2018) On the one hand OM is directlyassociated with carbon concentration in the soil one ofthe most stable carbon reservoirs worldwide thus large-scale losses of OM due to fire could have important impli-cations for global climate change (Lal 2004) On the otherhand nitrogen is often the most limiting soil nutrient forplant growth and its loss directly affects ecosystems by re-ducing net primary productivity (Vitousek and Howarth1991 Pellegrini et al 2018) In addition our results of ahigher negative effect of fire on OM in semiarid cli-mates compared to humid-warm climate is supportedby Pellegrini et al (2018) who found that across sa-vannas and grasslands biannual fire simulation induceshigh carbon losses under drier climates However ithas been recently proposed that a single fire event maybe related to either an increase a decrease or nochange in soil carbon and nitrogen while frequent firesover time produce a consistent decrease of both ele-ments across ecosystems (Pellegrini et al 2020a b)This hypothesis poses a constraint to our analysessince most of the articles did not report fire history fortheir study sites Consequently the heterogeneous ef-fects of single fires on soil carbon and nitrogen may beconfused with those of frequent fires therefore limitingthe identification of clearer patternsFire produces an increase in bulk density as a result

of soil disaggregation due to the decrease or loss ofOM and to the infiltration of ashes into soil micro-pores (Boyer and Miller 1994 Mataix-Solera et al2011) Soil pH also increased in burned plots as aresult of organic acid denaturation due to soil heatingand the accumulation of potassium and sodiumhydroxides and magnesium and calcium carbonates

Giorgis et al Fire Ecology (2021) 1711 Page 13 of 20

(Knicker 2007) The increase in pH is also driven bythe reduction in OM (Certini 2005) Additionallyimmediate phosphorus enrichment after a fire event isthe result of organic pool conversion of phosphorusinto orthophosphate which declines at later post-firetimes as it is also suggested by Certini (2005)Climatendashfirendashvegetation feedbacks Our review sug-

gests that climatic conditions associated with higherrates of biomass accumulation might determine a fasterecosystem recovery in burned sites across South Amer-ica We found a faster recovery in invertebrate abun-dance and lower losses of soil organic matter underclimates with higher water availability and temperaturesthan in semiarid climates Beyond climatic conditions inhistorically fire-prone ecosystems where fire has pre-sumably been an important evolutionary factor (Bondand Keeley 2005) a greater recovery of the biota is ex-pected compared to non-fire-prone ecosystems Howeverour classification of climate types may not necessarily re-flect the fire history of the corresponding regions there-fore making a clear interpretation of the reported patternsdifficult Although semiarid climate sites could be clearlyconsidered as fire-prone systems (Bond 2005) warm-humid climate sites may also include fire-prone ecosys-tems because some of them have seasonal rainfall pat-terns which is a key characteristic in determining the fireproneness of an ecosystem (Nogueira et al 2017 Romanoand Ursino 2020) In fact following the classification ofOlson et al (2001) 40 of the studies from warm-humidareas corresponded to tropical and subtropical grasslandsavanna and shrublands biome (arguably fire prone)while the other 60 corresponded to tropical and sub-tropical moist broadleaf forest biome (arguably non-fireprone) Most of the former were located in the Cerradoregion where Hoffmann and Moreira (2002) found a fas-ter post-fire recovery of woody species compared toneighboring tropical forest areas Thus they concludedthat the combination of fire-prone vegetation presumablywith a long fire history and warm-humid climate mightfavor a fast ecosystem recovery after fire Therefore theobserved results of greater recovery in warm-humid eco-systems types is likely due to a combination of both cli-mate and fire history Likewise most of the studies fromcold-humid areas have a mediterranean-type climate (egnorthwest Patagonia in Argentina Keeley et al 2012)where rainy winters and hot and dry summers make theseregions fire prone Such regions have been historicallysubjected to periodic fires (both natural and anthropo-genic) and therefore many plant species are capable of re-covering after a fire event (Veblen et al 2003 2008Defosseacute et al 2015) Interestingly we found a significantpost-fire decrease in nitrogen content and soil OM undersemiarid climate but not under cold-humid climateThis response pattern may be due to the inherent

differences in ecosystem properties between climatetypes (eg decomposition and growth rates succes-sional patterns and fire regimen among othersMcLauchlan et al 2020) Fire regime might play animportant role as more frequent surface fire in semi-arid ecosystems could lead to a lower recovery of soilproperties compared to cold-humid ecosystems wherecrown fires predominate and fire frequency is lower thanin semiarid ecosystems therefore buffering fire effects onsoil propertiesThe lack of a standardized classification of fire-

prone and non-fire-prone regions as well as the lim-ited available information on the frequency intensityand severity of fires in the research studies includedin this review are certainly limitations that call forprecaution in the interpretation and extrapolation ofthe results reported here Overcoming such limita-tions remains a challenge in fire ecology (Harris et al2016 Kelly et al 2018) We believe that our studyrepresents a step forward in the synthesis of fire-related patterns at a large scale Our meta-analysis aswell as the discussed patterns provides relevant infor-mation toward understanding the feedback betweenclimate vegetation and fire on the South Americancontinent An ultimately desirable goal in the mediumor long term would be the development of a globalnetwork of fire field experimental surveys across sys-tems with different vegetation types and fire histories

Future perspectivesAcross recently published global fire reviews there isclearly a low representation of studies from SouthAmerica (Prichard et al 2017 Geary et al 2019)Given the relevance of this continent in the globaldynamics of fire it is highly necessary that morestudies on different aspects of fire ecology be carriedout throughout fire-prone regions of the continent(eg seasonal forests and humid and semiarid savannas)Furthermore our review allowed us to identify specific re-search gaps which can be added to the agenda of researchpriorities for future studies in fire ecology (McLauchlanet al 2020)Most of the analyzed studies failed to include fire

characteristics such as frequency intensity and severitywhich prevented us from obtaining a reliable pattern oftheir role as modulators of ecosystem and organismresponses to fire This is an unmet challenge in fireresearch as these fire characteristics may certainly influ-ence the trends reported in this review (Keeley 2009Fontaine and Kennedy 2012 Balch et al 2015 Silveiraet al 2016 Carbone et al 2019)Considering that woody ecosystems might take more

than 40 years to recover after fire (Cavallero et al 2015

Giorgis et al Fire Ecology (2021) 1711 Page 14 of 20

Doherty et al 2017) studies on fire responses that spana short recovery period fail to elucidate what might hap-pen in the meantime between fire occurrence and recov-ery Consequently our review poses an urgent call toincrease the studies assessing changes in fire effects overlonger post-fire times than those usually consideredFurther studies focusing on biodiversity are needed to

increase sensibility to detect changes across taxa groupsclimate types and post-fire timesAdditionally there is a gap of studies on vertebrate

responses to fire and some of the few available ones havemethodological limitations which prevented us from in-cluding them in our analysis (eg lack of replicates inmammals studies Griffiths and Brook 2014)Studies dealing with fire and exotic or invasive species

were quite scarce While invasion ecology has shown in-creasing attention in the last decades (eg Gurevitch et al2011 Pyšek et al 2020) it does not seem to be reflected infire studies in South America (but see Chaneton et al2004 Hoffmann et al 2004 Maziacutea et al 2010 Raffaeleet al 2016 Herrero et al 2016 Marcora et al 2018) Webelieve that this should be a priority topic for future re-search particularly considering the tight link betweenplant invasions and changes in fire regimes (Brooks et al2004 Mandle et al 2011 Harris et al 2016)

Supplementary InformationThe online version contains supplementary material available at httpsdoiorg101186s42408-021-00100-9

Additional file 1 Supplementary material regarding the choice ofstudies published between 1990 and 2019 that were used in the reviewof fire effects across South American ecosystems to determine the role ofclimate and time since fire Figure 11 PRISMA flow diagramrepresenting the flow of information through the decision process (iethe number of studies identified rejected and accepted) for the meta-analysis of fire effects across South American ecosystems to determinethe role of climate and time since fire Studies used were published be-tween 1990 and 2019 List 11 List of reviewed studies published be-tween 1990 and 2019 that were used in the meta-analysis of fire effectsacross South American ecosystems to determine the role of climate andtime since fire

Additional file 2 Distribution across the climatic temperaturendashprecipitation plane of study sites (n = 175) from the 160 articles reviewedfor the meta-analysis of fire effects across South American ecosystems todetermine the role of climate and time since fire Studies used were pub-lished between 1990 and 2019 The four climate types established for thecollected data are indicated by different colors

Additional file 3 Distribution across the climatic temperaturendashprecipitation plane of study sites (n = 175) from the 160 articles reviewedfor the meta-analysis of fire effects across South American ecosystems todetermine the role of climate and time since fire Studies used were pub-lished between 1990 and 2019 The four climate types established for thecollected data are indicated by different colors Table 31 Synopsis ofmodels used in the different hierarchical mixed effects meta-analyses offire effects across South American ecosystems taken from studies pub-lished between 1990 and 2019 to determine the role of climate and timesince fire expressed in R language syntax using the metafor package(Viechtbauer 2010) yi = Hedgesrsquo d vi = variance of Hedgesrsquo d ID = effectidentity Table 32 Summary of the fire effects on biodiversity abun-dance fitness and soil properties showing tests of moderator variables

and heterogeneities from the meta-analyses of fire effects across SouthAmerican ecosystems to determine the role of climate and time sincefire Studies used were published between 1990 and 2019 Significant in-fluence of moderators is indicated in bold when P le 005 and the 95confidence interval (CI) does not include zero Non-significant effects (P gt005) are indicated with ldquonsrdquo Note that Q-total (QT) statistics are reportedfor overall effect and Q-between (Qb) statistics are reported for categor-ical moderator variables

Additional file 4 Supplementary information 41 Publication biasfor the meta-analysis of fire effects across South American ecosystems todetermine the role of climate and time since fire Studies used were pub-lished between 1990 and 2019 Table 41 Sample size (number of obser-vations) and results of publication bias and hierarchical models testingfor each response variable in our meta-analysis of fire effects across SouthAmerican ecosystems to determine the role of climate and time sincefire Studies used were published between 1990 and 2019 Publicationbias was tested using fail-safe number (ie Rosenberg method) Kendallrsquosrank correlation test (z) and ldquotrim and fillrdquo models

AcknowledgementsMAG and HvW were partially supported by the National Scientific andTechnological Research Council (CONICET) fellow program MAG SRZ LC RAAF PAT EK FB DEG PV and PJ are researchers of CONICET Argentina MAGLC PAT EK FB and DEG are professors at the National University of Cordobaand PV is professor at the Universidad Nacional de Cuyo

Authorsrsquo contributionsMAG SRZ LC RA HZ HvW and PJ designed the research and analysesMAG SRZ LC RA and PJ conducted literature search with the contributionof all the authors MAG LC and PJ wrote the article with contributions fromRA AF PT and EK All authors read and approved the final manuscript

FundingThe National Council for Scientific and Technological Research (CONICET) ofArgentina indirectly supported this manuscript through exchange grants forMAG and HvW However no direct funding supported this project

Availability of data and materialsNo unique data were created in the writing of this manuscript so noadditional data are available

Declarations

Ethics approval and consent to participateNot applicable

Consent for publicationConsent for publication not applicable as we use previously published

Competing interestsThe authors declare that they have no competing interests

Author details1Instituto Multidisciplinario de Biologia Vegetal (CONICET-UNC) VelezSarsfield 1611 CC 495 5000 Cordoba Argentina 2Facultad de CienciasExactas Fisicas y Naturales Universidad Nacional de Cordoba Velez Sarsfield299 CC 495 5000 Cordoba Argentina 3Faculty of Sustainability LeuphanaUniversity Luneburg Institute for Ethics and Transdisciplinary SustainabilityResearch Universitatsallee 1 21335 Luneburg Germany 4Faculty ofSustainability Leuphana University Luneburg Institute of EcologyUniversitatsallee 1 21335 Luneburg Germany 5Instituto de Diversidad yEcologia Animal (IDEA) CONICET-UNC and Facultad de Ciencias ExactasFisicas y Naturales Universidad Nacional de Cordoba Velez Sarsfield 299 CP5000 Cordoba Argentina 6Facultad de Ciencias Agrarias UniversidadNacional de Cuyo Brown sin nuacutemero Chacras de Coria 5505 MendozaArgentina 7Facultad de Ciencias Agrarias Universidad Nacional de CuyoM5528AHB Chacras de Coria Mendoza Argentina

Giorgis et al Fire Ecology (2021) 1711 Page 15 of 20

Fig 5 Weighted-mean effect sizes and 95 bias-corrected confidence intervals of fire on abundance from studies published between 1990 and 2019that were used in the review of fire effects across South American ecosystems to determine the role of climate and time since fire The effect sizes ofearly (-early) and late (-late) post-fire time for abundance of (A) plants invertebrates and vertebrates and (B) grasses shrubs and trees (C) The effectsizes of shrubs trees invertebrates and vertebrates across different climate types (-dry -semiarid -humid-warm) Parameters with confidence intervalsthat do not overlap the vertical dotted line (Hedgersquos d = 0) are considered to have a significant positive or negative effect Sample sizes for eachcategory are shown in parentheses The size of each black dot is proportional to its weight or contribution to the overall mean calculation None ofthe effect sizes of the groups within each moderator variable are significantly different (see Additional file 3 Table 32 for heterogeneity tests)

Giorgis et al Fire Ecology (2021) 1711 Page 10 of 20

growth forms early and late post fire (Fig 5B) is expectedconsidering that the relative recovery of pre-fire abun-dance (including cover and biomass) is slower than inherbaceous species The strongest negative effect observedfor shrubs could be mainly due to the fact that most ofthe studies were carried out in dry-climate ecosystemswhere the combination of water stress and fire signifi-cantly limits shrub recovery (Casillo et al 2012 Pratt et al2014 Jacobsen et al 2016)After fire a combination of different processes may

limit woody species recovery First woody species innon-fire-prone ecosystems may experience high mortal-ity rates (Barlow et al 2003) Second the reduction inaboveground biomass produced by fires exposes individ-uals to direct influence of browsing and trampling bylivestock (Coop et al 2010 Blackhall et al 2015 2017Marcora et al 2018 OrsquoConnor et al 2020 Zeballos et al2020) Third recruitment of woody species seedlingsinto fire-promoted grasslands or herbaceous patchesmay be more restricted due to an increase in competi-tion for light water or nutrients (Scholes and Archer1997 Casillo et al 2012) Finally soil erosion driven byfire increases run-off and seed loss events and reducesnutrient availability therefore constraining biomass pro-duction (Silva et al 2013 Balch et al 2015) The lessernegative effect observed in forbs and grasses may be at-tributed to their relative higher growth rate and shorterlifespan as compared to woody species which result in afaster recovery in burned habitats This vegetativeresponse might be linked to the responses observed insome of the mobile organisms (Fig 4) For example theearly recovery of herbaceous vegetation could promoterapid re-colonization by invertebrates such as insectherbivores and pollinators Additionally our resultsindicated different trends of abundance recovery acrosstaxa groups of invertebrates and vertebrates Similarly aprevious review showed different patterns of ldquooptimumrdquoabundance as time since fire increased (Doherty et al2017) The fast recovery of invertebrate abundance earlypost fire was mainly driven by insect studies (Carboneet al 2019) particularly on Orthoptera (Silveira et al2010 Kral et al 2017) In turn the overall negative effectof fire on the abundance of vertebrates was mainlydriven by bird studies which mostly showed a negativeresponse to fire (Fig 5) Although this pattern has beenobserved in other ecosystems (Fontaine and Kennedy2012 Doherty et al 2017 Carbone et al 2019) the re-sponse of birds to fire can be heterogeneous varyingconsiderably between taxonomic and functional groups(Fontaine and Kennedy 2012) therefore generalizationsmay be difficult to establish The pattern observed herecould be due to the fact that the systematically selectedstudies focused on rather intense fires or on bird speciesthat are more susceptible to fire Our findings of some

Fig 6 Weighted-mean effect sizes and 95 bias-corrected confidenceintervals of fire on plant fitness from studies published between 1990and 2019 that were used in the review of fire effects across SouthAmerican ecosystems to determine the role of climate and time sincefire (A) Effect sizes within growth forms and early and late post-firetime (B) Effect sizes of vegetative and reproductive fitness withingrowth forms (C) Effect sizes of vegetative (-veget) and reproductive(-reprod) fitness within growth forms and early (-early) and late (-late)post-fire time Parameters with confidence intervals that do notoverlap the vertical dotted line (Hedgersquos d = 0) are considered to havea significant positive or negative effect Sample sizes for each categoryare shown in parentheses The size of each black dot is proportional toits weight or contribution to the overall mean calculation Asterisks ()denotes a significant difference (Q-between [Qb] statistics) amongcategories ( = P lt 00001 = P lt 005)

Giorgis et al Fire Ecology (2021) 1711 Page 11 of 20

level of recovery of vertebrate abundance late post firecompared to early post fire (Fig 5) could have beendriven by the re-establishment of some bird populations

over time (Fontaine and Kennedy 2012 Doherty et al2017) Additionally the recovery of populations ofspecialist birds may take longer particularly in forest

Fig 7 Weighted-mean effect sizes and 95 bias-corrected confidence intervals of fire on soil properties from studies published between 1990and 2019 that were used in the review of fire effects across South American ecosystems to determine the role of climate and time since fire (A)Effect sizes of bulk density litter organic matter (OM) total nitrogen (Total N) available nitrogen (N) phosphorous (P) micronutrients pH andsalinity are shown (B) Effect sizes of early (-early) and late (-late) post-fire time of organic matter total nitrogen available nitrogen phosphorousand within climate type (-humid-warm -humid-cold -semiarid) of organic matter and total nitrogen Parameters with confidence intervals that donot overlap the vertical dotted line (Hedgersquos d = 0) are considered to have a significant positive or negative effect Sample sizes for each categoryare shown in parentheses The size of each black dot is proportional to its weight or contribution to the overall mean calculation Asterisk ()denotes a significant difference (Q-between [Qb] statistics) among categories ( = P lt 005)

Giorgis et al Fire Ecology (2021) 1711 Page 12 of 20

ecosystems (eg Barlow and Peres 2004 Fontaine andKennedy 2012 Mestre et al 2013 Albanesi et al 2014)although there are exceptions (eg Fontaine andKennedy 2012 Lee 2018 Morales et al 2020) such asthe group of birds known as cavity excavators (iewoodpeckers [Picidae]) that take advantage of both thesoftness of the wood of recently burned trees (ie snags)for making their cavities (Schepps et al 1999 Winklerand Christie 2002 Bond et al 2012 Lorenz et al 2015)and the higher amount of food (insect and other inverte-brates) found in that substrate right after the fire(Murphy and Lehnhausen 1998 Nappi et al 2003)Plant fitness Despite the negative effect of fire on tree

abundance we found no effect of fire on tree fitnessand a slight positive trend on vegetative fitness early postfire (Fig 6A B) Many tree species in the analyzed eco-systems are highly resilient to individual fires mainlythrough resprouting (Hoffmann and Solbrig 2003 Sou-chie et al 2017 Jaureguiberry et al 2020) therefore ifburned plots do not burn again trees might grow andrecover their biomass (Doherty et al 2017 Miller et al2019 Coop et al 2020) This response does not neces-sarily imply a direct adaptation to fire as the fire-adaptivetraits on which it relies could also have arisen in responseto other factors that have had a long history in some ofthe studied regions such as herbivory and seasonaldrought (Keeley et al 2011) which indirectly provide thevegetation with a high resilience to fire On the otherhand the negative tendency of the reproductive fitness ob-served early post fire is expected since trees may take along time to recover their reproductive size (Chapin IIIet al 2011) In contrast to trees shrubs did show an in-crease in vegetative and reproductive fitness early post firewhich differs significantly from that of late post fire(Fig 6C) The early vegetative response reflects the highresprouting capacity of many shrub species across thestudied regions especially in fire-prone regions (Gurvichet al 2005 Bravo et al 2014 Torres et al 2014 Duriganet al 2020 Jaureguiberry et al 2020) Regarding the repro-ductive response the pattern reported here agrees withprevious evidence suggesting that shrubs might reachminimum reproductive sizes soon after fire then slowdown their growth rate and reproduction as successionprogresses (Hoffmann and Moreira 2002 Hoffmann andSolbrig 2003 Galiacutendez et al 2009) In the case of herb-aceous plants surprisingly we did not find a general effectin fitness but a positive trend in reproductive fitness wasobserved mainly early post fire for grasses (Fig 6B C)The lack of a consistent general response pattern in forbscould be explained by the high heterogeneity of functionalgroups included within this growth form (eg annual per-ennial climber fern) that can have different responses tofire (Keeley et al 1981 Bates et al 2014 Heydari et al2016 Arcamone and Jaureguiberry 2018 Vidaller et al

2019) In the case of grasses the tendency of increasing re-productive fitness early post fire (Fig 6C) could be relatedto the positive firendashcolonization feedback generally de-scribed for this growth form (Bond et al 2003 Bond andKeeley 2005 Pausas and Keeley 2009 Pilon et al 2018)Previous studies had observed an increase in reproductivefitness of grasses after fire (Baruch and Bilbao 1999 Ara-uacutejo et al 2013 Pilon et al 2018 Vidaller et al 2019) Onthe contrary vegetative fitness showed a negative impactearly post fire However we cannot determine if the ob-served pattern (ie positive reproductive fitness versusnegative vegetative fitness) was due to intrinsic differencesbetween types of fitness or to differences in the type of cli-mate since all studies on reproductive fitness were lo-cated in humid-warm climate while vegetative fitnesswas mostly studied in semiarid and arid climatesSoil properties The responses of soil properties to fire in

South American ecosystems support the trend suggestedin previous global reviews (Certini 2005 Pellegrini et al2018) Fire significantly reduces the amount of two im-portant soil properties OM and nitrogen content (Pelle-grini et al 2018) On the one hand OM is directlyassociated with carbon concentration in the soil one ofthe most stable carbon reservoirs worldwide thus large-scale losses of OM due to fire could have important impli-cations for global climate change (Lal 2004) On the otherhand nitrogen is often the most limiting soil nutrient forplant growth and its loss directly affects ecosystems by re-ducing net primary productivity (Vitousek and Howarth1991 Pellegrini et al 2018) In addition our results of ahigher negative effect of fire on OM in semiarid cli-mates compared to humid-warm climate is supportedby Pellegrini et al (2018) who found that across sa-vannas and grasslands biannual fire simulation induceshigh carbon losses under drier climates However ithas been recently proposed that a single fire event maybe related to either an increase a decrease or nochange in soil carbon and nitrogen while frequent firesover time produce a consistent decrease of both ele-ments across ecosystems (Pellegrini et al 2020a b)This hypothesis poses a constraint to our analysessince most of the articles did not report fire history fortheir study sites Consequently the heterogeneous ef-fects of single fires on soil carbon and nitrogen may beconfused with those of frequent fires therefore limitingthe identification of clearer patternsFire produces an increase in bulk density as a result

of soil disaggregation due to the decrease or loss ofOM and to the infiltration of ashes into soil micro-pores (Boyer and Miller 1994 Mataix-Solera et al2011) Soil pH also increased in burned plots as aresult of organic acid denaturation due to soil heatingand the accumulation of potassium and sodiumhydroxides and magnesium and calcium carbonates

Giorgis et al Fire Ecology (2021) 1711 Page 13 of 20

(Knicker 2007) The increase in pH is also driven bythe reduction in OM (Certini 2005) Additionallyimmediate phosphorus enrichment after a fire event isthe result of organic pool conversion of phosphorusinto orthophosphate which declines at later post-firetimes as it is also suggested by Certini (2005)Climatendashfirendashvegetation feedbacks Our review sug-

gests that climatic conditions associated with higherrates of biomass accumulation might determine a fasterecosystem recovery in burned sites across South Amer-ica We found a faster recovery in invertebrate abun-dance and lower losses of soil organic matter underclimates with higher water availability and temperaturesthan in semiarid climates Beyond climatic conditions inhistorically fire-prone ecosystems where fire has pre-sumably been an important evolutionary factor (Bondand Keeley 2005) a greater recovery of the biota is ex-pected compared to non-fire-prone ecosystems Howeverour classification of climate types may not necessarily re-flect the fire history of the corresponding regions there-fore making a clear interpretation of the reported patternsdifficult Although semiarid climate sites could be clearlyconsidered as fire-prone systems (Bond 2005) warm-humid climate sites may also include fire-prone ecosys-tems because some of them have seasonal rainfall pat-terns which is a key characteristic in determining the fireproneness of an ecosystem (Nogueira et al 2017 Romanoand Ursino 2020) In fact following the classification ofOlson et al (2001) 40 of the studies from warm-humidareas corresponded to tropical and subtropical grasslandsavanna and shrublands biome (arguably fire prone)while the other 60 corresponded to tropical and sub-tropical moist broadleaf forest biome (arguably non-fireprone) Most of the former were located in the Cerradoregion where Hoffmann and Moreira (2002) found a fas-ter post-fire recovery of woody species compared toneighboring tropical forest areas Thus they concludedthat the combination of fire-prone vegetation presumablywith a long fire history and warm-humid climate mightfavor a fast ecosystem recovery after fire Therefore theobserved results of greater recovery in warm-humid eco-systems types is likely due to a combination of both cli-mate and fire history Likewise most of the studies fromcold-humid areas have a mediterranean-type climate (egnorthwest Patagonia in Argentina Keeley et al 2012)where rainy winters and hot and dry summers make theseregions fire prone Such regions have been historicallysubjected to periodic fires (both natural and anthropo-genic) and therefore many plant species are capable of re-covering after a fire event (Veblen et al 2003 2008Defosseacute et al 2015) Interestingly we found a significantpost-fire decrease in nitrogen content and soil OM undersemiarid climate but not under cold-humid climateThis response pattern may be due to the inherent

differences in ecosystem properties between climatetypes (eg decomposition and growth rates succes-sional patterns and fire regimen among othersMcLauchlan et al 2020) Fire regime might play animportant role as more frequent surface fire in semi-arid ecosystems could lead to a lower recovery of soilproperties compared to cold-humid ecosystems wherecrown fires predominate and fire frequency is lower thanin semiarid ecosystems therefore buffering fire effects onsoil propertiesThe lack of a standardized classification of fire-

prone and non-fire-prone regions as well as the lim-ited available information on the frequency intensityand severity of fires in the research studies includedin this review are certainly limitations that call forprecaution in the interpretation and extrapolation ofthe results reported here Overcoming such limita-tions remains a challenge in fire ecology (Harris et al2016 Kelly et al 2018) We believe that our studyrepresents a step forward in the synthesis of fire-related patterns at a large scale Our meta-analysis aswell as the discussed patterns provides relevant infor-mation toward understanding the feedback betweenclimate vegetation and fire on the South Americancontinent An ultimately desirable goal in the mediumor long term would be the development of a globalnetwork of fire field experimental surveys across sys-tems with different vegetation types and fire histories

Future perspectivesAcross recently published global fire reviews there isclearly a low representation of studies from SouthAmerica (Prichard et al 2017 Geary et al 2019)Given the relevance of this continent in the globaldynamics of fire it is highly necessary that morestudies on different aspects of fire ecology be carriedout throughout fire-prone regions of the continent(eg seasonal forests and humid and semiarid savannas)Furthermore our review allowed us to identify specific re-search gaps which can be added to the agenda of researchpriorities for future studies in fire ecology (McLauchlanet al 2020)Most of the analyzed studies failed to include fire

characteristics such as frequency intensity and severitywhich prevented us from obtaining a reliable pattern oftheir role as modulators of ecosystem and organismresponses to fire This is an unmet challenge in fireresearch as these fire characteristics may certainly influ-ence the trends reported in this review (Keeley 2009Fontaine and Kennedy 2012 Balch et al 2015 Silveiraet al 2016 Carbone et al 2019)Considering that woody ecosystems might take more

than 40 years to recover after fire (Cavallero et al 2015

Giorgis et al Fire Ecology (2021) 1711 Page 14 of 20

Doherty et al 2017) studies on fire responses that spana short recovery period fail to elucidate what might hap-pen in the meantime between fire occurrence and recov-ery Consequently our review poses an urgent call toincrease the studies assessing changes in fire effects overlonger post-fire times than those usually consideredFurther studies focusing on biodiversity are needed to

increase sensibility to detect changes across taxa groupsclimate types and post-fire timesAdditionally there is a gap of studies on vertebrate

responses to fire and some of the few available ones havemethodological limitations which prevented us from in-cluding them in our analysis (eg lack of replicates inmammals studies Griffiths and Brook 2014)Studies dealing with fire and exotic or invasive species

were quite scarce While invasion ecology has shown in-creasing attention in the last decades (eg Gurevitch et al2011 Pyšek et al 2020) it does not seem to be reflected infire studies in South America (but see Chaneton et al2004 Hoffmann et al 2004 Maziacutea et al 2010 Raffaeleet al 2016 Herrero et al 2016 Marcora et al 2018) Webelieve that this should be a priority topic for future re-search particularly considering the tight link betweenplant invasions and changes in fire regimes (Brooks et al2004 Mandle et al 2011 Harris et al 2016)

Supplementary InformationThe online version contains supplementary material available at httpsdoiorg101186s42408-021-00100-9

Additional file 1 Supplementary material regarding the choice ofstudies published between 1990 and 2019 that were used in the reviewof fire effects across South American ecosystems to determine the role ofclimate and time since fire Figure 11 PRISMA flow diagramrepresenting the flow of information through the decision process (iethe number of studies identified rejected and accepted) for the meta-analysis of fire effects across South American ecosystems to determinethe role of climate and time since fire Studies used were published be-tween 1990 and 2019 List 11 List of reviewed studies published be-tween 1990 and 2019 that were used in the meta-analysis of fire effectsacross South American ecosystems to determine the role of climate andtime since fire

Additional file 2 Distribution across the climatic temperaturendashprecipitation plane of study sites (n = 175) from the 160 articles reviewedfor the meta-analysis of fire effects across South American ecosystems todetermine the role of climate and time since fire Studies used were pub-lished between 1990 and 2019 The four climate types established for thecollected data are indicated by different colors

Additional file 3 Distribution across the climatic temperaturendashprecipitation plane of study sites (n = 175) from the 160 articles reviewedfor the meta-analysis of fire effects across South American ecosystems todetermine the role of climate and time since fire Studies used were pub-lished between 1990 and 2019 The four climate types established for thecollected data are indicated by different colors Table 31 Synopsis ofmodels used in the different hierarchical mixed effects meta-analyses offire effects across South American ecosystems taken from studies pub-lished between 1990 and 2019 to determine the role of climate and timesince fire expressed in R language syntax using the metafor package(Viechtbauer 2010) yi = Hedgesrsquo d vi = variance of Hedgesrsquo d ID = effectidentity Table 32 Summary of the fire effects on biodiversity abun-dance fitness and soil properties showing tests of moderator variables

and heterogeneities from the meta-analyses of fire effects across SouthAmerican ecosystems to determine the role of climate and time sincefire Studies used were published between 1990 and 2019 Significant in-fluence of moderators is indicated in bold when P le 005 and the 95confidence interval (CI) does not include zero Non-significant effects (P gt005) are indicated with ldquonsrdquo Note that Q-total (QT) statistics are reportedfor overall effect and Q-between (Qb) statistics are reported for categor-ical moderator variables

Additional file 4 Supplementary information 41 Publication biasfor the meta-analysis of fire effects across South American ecosystems todetermine the role of climate and time since fire Studies used were pub-lished between 1990 and 2019 Table 41 Sample size (number of obser-vations) and results of publication bias and hierarchical models testingfor each response variable in our meta-analysis of fire effects across SouthAmerican ecosystems to determine the role of climate and time sincefire Studies used were published between 1990 and 2019 Publicationbias was tested using fail-safe number (ie Rosenberg method) Kendallrsquosrank correlation test (z) and ldquotrim and fillrdquo models

AcknowledgementsMAG and HvW were partially supported by the National Scientific andTechnological Research Council (CONICET) fellow program MAG SRZ LC RAAF PAT EK FB DEG PV and PJ are researchers of CONICET Argentina MAGLC PAT EK FB and DEG are professors at the National University of Cordobaand PV is professor at the Universidad Nacional de Cuyo

Authorsrsquo contributionsMAG SRZ LC RA HZ HvW and PJ designed the research and analysesMAG SRZ LC RA and PJ conducted literature search with the contributionof all the authors MAG LC and PJ wrote the article with contributions fromRA AF PT and EK All authors read and approved the final manuscript

FundingThe National Council for Scientific and Technological Research (CONICET) ofArgentina indirectly supported this manuscript through exchange grants forMAG and HvW However no direct funding supported this project

Availability of data and materialsNo unique data were created in the writing of this manuscript so noadditional data are available

Declarations

Ethics approval and consent to participateNot applicable

Consent for publicationConsent for publication not applicable as we use previously published

Competing interestsThe authors declare that they have no competing interests

Author details1Instituto Multidisciplinario de Biologia Vegetal (CONICET-UNC) VelezSarsfield 1611 CC 495 5000 Cordoba Argentina 2Facultad de CienciasExactas Fisicas y Naturales Universidad Nacional de Cordoba Velez Sarsfield299 CC 495 5000 Cordoba Argentina 3Faculty of Sustainability LeuphanaUniversity Luneburg Institute for Ethics and Transdisciplinary SustainabilityResearch Universitatsallee 1 21335 Luneburg Germany 4Faculty ofSustainability Leuphana University Luneburg Institute of EcologyUniversitatsallee 1 21335 Luneburg Germany 5Instituto de Diversidad yEcologia Animal (IDEA) CONICET-UNC and Facultad de Ciencias ExactasFisicas y Naturales Universidad Nacional de Cordoba Velez Sarsfield 299 CP5000 Cordoba Argentina 6Facultad de Ciencias Agrarias UniversidadNacional de Cuyo Brown sin nuacutemero Chacras de Coria 5505 MendozaArgentina 7Facultad de Ciencias Agrarias Universidad Nacional de CuyoM5528AHB Chacras de Coria Mendoza Argentina

Giorgis et al Fire Ecology (2021) 1711 Page 15 of 20

growth forms early and late post fire (Fig 5B) is expectedconsidering that the relative recovery of pre-fire abun-dance (including cover and biomass) is slower than inherbaceous species The strongest negative effect observedfor shrubs could be mainly due to the fact that most ofthe studies were carried out in dry-climate ecosystemswhere the combination of water stress and fire signifi-cantly limits shrub recovery (Casillo et al 2012 Pratt et al2014 Jacobsen et al 2016)After fire a combination of different processes may

limit woody species recovery First woody species innon-fire-prone ecosystems may experience high mortal-ity rates (Barlow et al 2003) Second the reduction inaboveground biomass produced by fires exposes individ-uals to direct influence of browsing and trampling bylivestock (Coop et al 2010 Blackhall et al 2015 2017Marcora et al 2018 OrsquoConnor et al 2020 Zeballos et al2020) Third recruitment of woody species seedlingsinto fire-promoted grasslands or herbaceous patchesmay be more restricted due to an increase in competi-tion for light water or nutrients (Scholes and Archer1997 Casillo et al 2012) Finally soil erosion driven byfire increases run-off and seed loss events and reducesnutrient availability therefore constraining biomass pro-duction (Silva et al 2013 Balch et al 2015) The lessernegative effect observed in forbs and grasses may be at-tributed to their relative higher growth rate and shorterlifespan as compared to woody species which result in afaster recovery in burned habitats This vegetativeresponse might be linked to the responses observed insome of the mobile organisms (Fig 4) For example theearly recovery of herbaceous vegetation could promoterapid re-colonization by invertebrates such as insectherbivores and pollinators Additionally our resultsindicated different trends of abundance recovery acrosstaxa groups of invertebrates and vertebrates Similarly aprevious review showed different patterns of ldquooptimumrdquoabundance as time since fire increased (Doherty et al2017) The fast recovery of invertebrate abundance earlypost fire was mainly driven by insect studies (Carboneet al 2019) particularly on Orthoptera (Silveira et al2010 Kral et al 2017) In turn the overall negative effectof fire on the abundance of vertebrates was mainlydriven by bird studies which mostly showed a negativeresponse to fire (Fig 5) Although this pattern has beenobserved in other ecosystems (Fontaine and Kennedy2012 Doherty et al 2017 Carbone et al 2019) the re-sponse of birds to fire can be heterogeneous varyingconsiderably between taxonomic and functional groups(Fontaine and Kennedy 2012) therefore generalizationsmay be difficult to establish The pattern observed herecould be due to the fact that the systematically selectedstudies focused on rather intense fires or on bird speciesthat are more susceptible to fire Our findings of some

Fig 6 Weighted-mean effect sizes and 95 bias-corrected confidenceintervals of fire on plant fitness from studies published between 1990and 2019 that were used in the review of fire effects across SouthAmerican ecosystems to determine the role of climate and time sincefire (A) Effect sizes within growth forms and early and late post-firetime (B) Effect sizes of vegetative and reproductive fitness withingrowth forms (C) Effect sizes of vegetative (-veget) and reproductive(-reprod) fitness within growth forms and early (-early) and late (-late)post-fire time Parameters with confidence intervals that do notoverlap the vertical dotted line (Hedgersquos d = 0) are considered to havea significant positive or negative effect Sample sizes for each categoryare shown in parentheses The size of each black dot is proportional toits weight or contribution to the overall mean calculation Asterisks ()denotes a significant difference (Q-between [Qb] statistics) amongcategories ( = P lt 00001 = P lt 005)

Giorgis et al Fire Ecology (2021) 1711 Page 11 of 20

level of recovery of vertebrate abundance late post firecompared to early post fire (Fig 5) could have beendriven by the re-establishment of some bird populations

over time (Fontaine and Kennedy 2012 Doherty et al2017) Additionally the recovery of populations ofspecialist birds may take longer particularly in forest

Fig 7 Weighted-mean effect sizes and 95 bias-corrected confidence intervals of fire on soil properties from studies published between 1990and 2019 that were used in the review of fire effects across South American ecosystems to determine the role of climate and time since fire (A)Effect sizes of bulk density litter organic matter (OM) total nitrogen (Total N) available nitrogen (N) phosphorous (P) micronutrients pH andsalinity are shown (B) Effect sizes of early (-early) and late (-late) post-fire time of organic matter total nitrogen available nitrogen phosphorousand within climate type (-humid-warm -humid-cold -semiarid) of organic matter and total nitrogen Parameters with confidence intervals that donot overlap the vertical dotted line (Hedgersquos d = 0) are considered to have a significant positive or negative effect Sample sizes for each categoryare shown in parentheses The size of each black dot is proportional to its weight or contribution to the overall mean calculation Asterisk ()denotes a significant difference (Q-between [Qb] statistics) among categories ( = P lt 005)

Giorgis et al Fire Ecology (2021) 1711 Page 12 of 20

ecosystems (eg Barlow and Peres 2004 Fontaine andKennedy 2012 Mestre et al 2013 Albanesi et al 2014)although there are exceptions (eg Fontaine andKennedy 2012 Lee 2018 Morales et al 2020) such asthe group of birds known as cavity excavators (iewoodpeckers [Picidae]) that take advantage of both thesoftness of the wood of recently burned trees (ie snags)for making their cavities (Schepps et al 1999 Winklerand Christie 2002 Bond et al 2012 Lorenz et al 2015)and the higher amount of food (insect and other inverte-brates) found in that substrate right after the fire(Murphy and Lehnhausen 1998 Nappi et al 2003)Plant fitness Despite the negative effect of fire on tree

abundance we found no effect of fire on tree fitnessand a slight positive trend on vegetative fitness early postfire (Fig 6A B) Many tree species in the analyzed eco-systems are highly resilient to individual fires mainlythrough resprouting (Hoffmann and Solbrig 2003 Sou-chie et al 2017 Jaureguiberry et al 2020) therefore ifburned plots do not burn again trees might grow andrecover their biomass (Doherty et al 2017 Miller et al2019 Coop et al 2020) This response does not neces-sarily imply a direct adaptation to fire as the fire-adaptivetraits on which it relies could also have arisen in responseto other factors that have had a long history in some ofthe studied regions such as herbivory and seasonaldrought (Keeley et al 2011) which indirectly provide thevegetation with a high resilience to fire On the otherhand the negative tendency of the reproductive fitness ob-served early post fire is expected since trees may take along time to recover their reproductive size (Chapin IIIet al 2011) In contrast to trees shrubs did show an in-crease in vegetative and reproductive fitness early post firewhich differs significantly from that of late post fire(Fig 6C) The early vegetative response reflects the highresprouting capacity of many shrub species across thestudied regions especially in fire-prone regions (Gurvichet al 2005 Bravo et al 2014 Torres et al 2014 Duriganet al 2020 Jaureguiberry et al 2020) Regarding the repro-ductive response the pattern reported here agrees withprevious evidence suggesting that shrubs might reachminimum reproductive sizes soon after fire then slowdown their growth rate and reproduction as successionprogresses (Hoffmann and Moreira 2002 Hoffmann andSolbrig 2003 Galiacutendez et al 2009) In the case of herb-aceous plants surprisingly we did not find a general effectin fitness but a positive trend in reproductive fitness wasobserved mainly early post fire for grasses (Fig 6B C)The lack of a consistent general response pattern in forbscould be explained by the high heterogeneity of functionalgroups included within this growth form (eg annual per-ennial climber fern) that can have different responses tofire (Keeley et al 1981 Bates et al 2014 Heydari et al2016 Arcamone and Jaureguiberry 2018 Vidaller et al

2019) In the case of grasses the tendency of increasing re-productive fitness early post fire (Fig 6C) could be relatedto the positive firendashcolonization feedback generally de-scribed for this growth form (Bond et al 2003 Bond andKeeley 2005 Pausas and Keeley 2009 Pilon et al 2018)Previous studies had observed an increase in reproductivefitness of grasses after fire (Baruch and Bilbao 1999 Ara-uacutejo et al 2013 Pilon et al 2018 Vidaller et al 2019) Onthe contrary vegetative fitness showed a negative impactearly post fire However we cannot determine if the ob-served pattern (ie positive reproductive fitness versusnegative vegetative fitness) was due to intrinsic differencesbetween types of fitness or to differences in the type of cli-mate since all studies on reproductive fitness were lo-cated in humid-warm climate while vegetative fitnesswas mostly studied in semiarid and arid climatesSoil properties The responses of soil properties to fire in

South American ecosystems support the trend suggestedin previous global reviews (Certini 2005 Pellegrini et al2018) Fire significantly reduces the amount of two im-portant soil properties OM and nitrogen content (Pelle-grini et al 2018) On the one hand OM is directlyassociated with carbon concentration in the soil one ofthe most stable carbon reservoirs worldwide thus large-scale losses of OM due to fire could have important impli-cations for global climate change (Lal 2004) On the otherhand nitrogen is often the most limiting soil nutrient forplant growth and its loss directly affects ecosystems by re-ducing net primary productivity (Vitousek and Howarth1991 Pellegrini et al 2018) In addition our results of ahigher negative effect of fire on OM in semiarid cli-mates compared to humid-warm climate is supportedby Pellegrini et al (2018) who found that across sa-vannas and grasslands biannual fire simulation induceshigh carbon losses under drier climates However ithas been recently proposed that a single fire event maybe related to either an increase a decrease or nochange in soil carbon and nitrogen while frequent firesover time produce a consistent decrease of both ele-ments across ecosystems (Pellegrini et al 2020a b)This hypothesis poses a constraint to our analysessince most of the articles did not report fire history fortheir study sites Consequently the heterogeneous ef-fects of single fires on soil carbon and nitrogen may beconfused with those of frequent fires therefore limitingthe identification of clearer patternsFire produces an increase in bulk density as a result

of soil disaggregation due to the decrease or loss ofOM and to the infiltration of ashes into soil micro-pores (Boyer and Miller 1994 Mataix-Solera et al2011) Soil pH also increased in burned plots as aresult of organic acid denaturation due to soil heatingand the accumulation of potassium and sodiumhydroxides and magnesium and calcium carbonates

Giorgis et al Fire Ecology (2021) 1711 Page 13 of 20

(Knicker 2007) The increase in pH is also driven bythe reduction in OM (Certini 2005) Additionallyimmediate phosphorus enrichment after a fire event isthe result of organic pool conversion of phosphorusinto orthophosphate which declines at later post-firetimes as it is also suggested by Certini (2005)Climatendashfirendashvegetation feedbacks Our review sug-

gests that climatic conditions associated with higherrates of biomass accumulation might determine a fasterecosystem recovery in burned sites across South Amer-ica We found a faster recovery in invertebrate abun-dance and lower losses of soil organic matter underclimates with higher water availability and temperaturesthan in semiarid climates Beyond climatic conditions inhistorically fire-prone ecosystems where fire has pre-sumably been an important evolutionary factor (Bondand Keeley 2005) a greater recovery of the biota is ex-pected compared to non-fire-prone ecosystems Howeverour classification of climate types may not necessarily re-flect the fire history of the corresponding regions there-fore making a clear interpretation of the reported patternsdifficult Although semiarid climate sites could be clearlyconsidered as fire-prone systems (Bond 2005) warm-humid climate sites may also include fire-prone ecosys-tems because some of them have seasonal rainfall pat-terns which is a key characteristic in determining the fireproneness of an ecosystem (Nogueira et al 2017 Romanoand Ursino 2020) In fact following the classification ofOlson et al (2001) 40 of the studies from warm-humidareas corresponded to tropical and subtropical grasslandsavanna and shrublands biome (arguably fire prone)while the other 60 corresponded to tropical and sub-tropical moist broadleaf forest biome (arguably non-fireprone) Most of the former were located in the Cerradoregion where Hoffmann and Moreira (2002) found a fas-ter post-fire recovery of woody species compared toneighboring tropical forest areas Thus they concludedthat the combination of fire-prone vegetation presumablywith a long fire history and warm-humid climate mightfavor a fast ecosystem recovery after fire Therefore theobserved results of greater recovery in warm-humid eco-systems types is likely due to a combination of both cli-mate and fire history Likewise most of the studies fromcold-humid areas have a mediterranean-type climate (egnorthwest Patagonia in Argentina Keeley et al 2012)where rainy winters and hot and dry summers make theseregions fire prone Such regions have been historicallysubjected to periodic fires (both natural and anthropo-genic) and therefore many plant species are capable of re-covering after a fire event (Veblen et al 2003 2008Defosseacute et al 2015) Interestingly we found a significantpost-fire decrease in nitrogen content and soil OM undersemiarid climate but not under cold-humid climateThis response pattern may be due to the inherent

differences in ecosystem properties between climatetypes (eg decomposition and growth rates succes-sional patterns and fire regimen among othersMcLauchlan et al 2020) Fire regime might play animportant role as more frequent surface fire in semi-arid ecosystems could lead to a lower recovery of soilproperties compared to cold-humid ecosystems wherecrown fires predominate and fire frequency is lower thanin semiarid ecosystems therefore buffering fire effects onsoil propertiesThe lack of a standardized classification of fire-

prone and non-fire-prone regions as well as the lim-ited available information on the frequency intensityand severity of fires in the research studies includedin this review are certainly limitations that call forprecaution in the interpretation and extrapolation ofthe results reported here Overcoming such limita-tions remains a challenge in fire ecology (Harris et al2016 Kelly et al 2018) We believe that our studyrepresents a step forward in the synthesis of fire-related patterns at a large scale Our meta-analysis aswell as the discussed patterns provides relevant infor-mation toward understanding the feedback betweenclimate vegetation and fire on the South Americancontinent An ultimately desirable goal in the mediumor long term would be the development of a globalnetwork of fire field experimental surveys across sys-tems with different vegetation types and fire histories

Future perspectivesAcross recently published global fire reviews there isclearly a low representation of studies from SouthAmerica (Prichard et al 2017 Geary et al 2019)Given the relevance of this continent in the globaldynamics of fire it is highly necessary that morestudies on different aspects of fire ecology be carriedout throughout fire-prone regions of the continent(eg seasonal forests and humid and semiarid savannas)Furthermore our review allowed us to identify specific re-search gaps which can be added to the agenda of researchpriorities for future studies in fire ecology (McLauchlanet al 2020)Most of the analyzed studies failed to include fire

characteristics such as frequency intensity and severitywhich prevented us from obtaining a reliable pattern oftheir role as modulators of ecosystem and organismresponses to fire This is an unmet challenge in fireresearch as these fire characteristics may certainly influ-ence the trends reported in this review (Keeley 2009Fontaine and Kennedy 2012 Balch et al 2015 Silveiraet al 2016 Carbone et al 2019)Considering that woody ecosystems might take more

than 40 years to recover after fire (Cavallero et al 2015

Giorgis et al Fire Ecology (2021) 1711 Page 14 of 20

Doherty et al 2017) studies on fire responses that spana short recovery period fail to elucidate what might hap-pen in the meantime between fire occurrence and recov-ery Consequently our review poses an urgent call toincrease the studies assessing changes in fire effects overlonger post-fire times than those usually consideredFurther studies focusing on biodiversity are needed to

increase sensibility to detect changes across taxa groupsclimate types and post-fire timesAdditionally there is a gap of studies on vertebrate

responses to fire and some of the few available ones havemethodological limitations which prevented us from in-cluding them in our analysis (eg lack of replicates inmammals studies Griffiths and Brook 2014)Studies dealing with fire and exotic or invasive species

were quite scarce While invasion ecology has shown in-creasing attention in the last decades (eg Gurevitch et al2011 Pyšek et al 2020) it does not seem to be reflected infire studies in South America (but see Chaneton et al2004 Hoffmann et al 2004 Maziacutea et al 2010 Raffaeleet al 2016 Herrero et al 2016 Marcora et al 2018) Webelieve that this should be a priority topic for future re-search particularly considering the tight link betweenplant invasions and changes in fire regimes (Brooks et al2004 Mandle et al 2011 Harris et al 2016)

Supplementary InformationThe online version contains supplementary material available at httpsdoiorg101186s42408-021-00100-9

Additional file 1 Supplementary material regarding the choice ofstudies published between 1990 and 2019 that were used in the reviewof fire effects across South American ecosystems to determine the role ofclimate and time since fire Figure 11 PRISMA flow diagramrepresenting the flow of information through the decision process (iethe number of studies identified rejected and accepted) for the meta-analysis of fire effects across South American ecosystems to determinethe role of climate and time since fire Studies used were published be-tween 1990 and 2019 List 11 List of reviewed studies published be-tween 1990 and 2019 that were used in the meta-analysis of fire effectsacross South American ecosystems to determine the role of climate andtime since fire

Additional file 2 Distribution across the climatic temperaturendashprecipitation plane of study sites (n = 175) from the 160 articles reviewedfor the meta-analysis of fire effects across South American ecosystems todetermine the role of climate and time since fire Studies used were pub-lished between 1990 and 2019 The four climate types established for thecollected data are indicated by different colors

Additional file 3 Distribution across the climatic temperaturendashprecipitation plane of study sites (n = 175) from the 160 articles reviewedfor the meta-analysis of fire effects across South American ecosystems todetermine the role of climate and time since fire Studies used were pub-lished between 1990 and 2019 The four climate types established for thecollected data are indicated by different colors Table 31 Synopsis ofmodels used in the different hierarchical mixed effects meta-analyses offire effects across South American ecosystems taken from studies pub-lished between 1990 and 2019 to determine the role of climate and timesince fire expressed in R language syntax using the metafor package(Viechtbauer 2010) yi = Hedgesrsquo d vi = variance of Hedgesrsquo d ID = effectidentity Table 32 Summary of the fire effects on biodiversity abun-dance fitness and soil properties showing tests of moderator variables

and heterogeneities from the meta-analyses of fire effects across SouthAmerican ecosystems to determine the role of climate and time sincefire Studies used were published between 1990 and 2019 Significant in-fluence of moderators is indicated in bold when P le 005 and the 95confidence interval (CI) does not include zero Non-significant effects (P gt005) are indicated with ldquonsrdquo Note that Q-total (QT) statistics are reportedfor overall effect and Q-between (Qb) statistics are reported for categor-ical moderator variables

Additional file 4 Supplementary information 41 Publication biasfor the meta-analysis of fire effects across South American ecosystems todetermine the role of climate and time since fire Studies used were pub-lished between 1990 and 2019 Table 41 Sample size (number of obser-vations) and results of publication bias and hierarchical models testingfor each response variable in our meta-analysis of fire effects across SouthAmerican ecosystems to determine the role of climate and time sincefire Studies used were published between 1990 and 2019 Publicationbias was tested using fail-safe number (ie Rosenberg method) Kendallrsquosrank correlation test (z) and ldquotrim and fillrdquo models

AcknowledgementsMAG and HvW were partially supported by the National Scientific andTechnological Research Council (CONICET) fellow program MAG SRZ LC RAAF PAT EK FB DEG PV and PJ are researchers of CONICET Argentina MAGLC PAT EK FB and DEG are professors at the National University of Cordobaand PV is professor at the Universidad Nacional de Cuyo

Authorsrsquo contributionsMAG SRZ LC RA HZ HvW and PJ designed the research and analysesMAG SRZ LC RA and PJ conducted literature search with the contributionof all the authors MAG LC and PJ wrote the article with contributions fromRA AF PT and EK All authors read and approved the final manuscript

FundingThe National Council for Scientific and Technological Research (CONICET) ofArgentina indirectly supported this manuscript through exchange grants forMAG and HvW However no direct funding supported this project

Availability of data and materialsNo unique data were created in the writing of this manuscript so noadditional data are available

Declarations

Ethics approval and consent to participateNot applicable

Consent for publicationConsent for publication not applicable as we use previously published

Competing interestsThe authors declare that they have no competing interests

Author details1Instituto Multidisciplinario de Biologia Vegetal (CONICET-UNC) VelezSarsfield 1611 CC 495 5000 Cordoba Argentina 2Facultad de CienciasExactas Fisicas y Naturales Universidad Nacional de Cordoba Velez Sarsfield299 CC 495 5000 Cordoba Argentina 3Faculty of Sustainability LeuphanaUniversity Luneburg Institute for Ethics and Transdisciplinary SustainabilityResearch Universitatsallee 1 21335 Luneburg Germany 4Faculty ofSustainability Leuphana University Luneburg Institute of EcologyUniversitatsallee 1 21335 Luneburg Germany 5Instituto de Diversidad yEcologia Animal (IDEA) CONICET-UNC and Facultad de Ciencias ExactasFisicas y Naturales Universidad Nacional de Cordoba Velez Sarsfield 299 CP5000 Cordoba Argentina 6Facultad de Ciencias Agrarias UniversidadNacional de Cuyo Brown sin nuacutemero Chacras de Coria 5505 MendozaArgentina 7Facultad de Ciencias Agrarias Universidad Nacional de CuyoM5528AHB Chacras de Coria Mendoza Argentina

Giorgis et al Fire Ecology (2021) 1711 Page 15 of 20

level of recovery of vertebrate abundance late post firecompared to early post fire (Fig 5) could have beendriven by the re-establishment of some bird populations

over time (Fontaine and Kennedy 2012 Doherty et al2017) Additionally the recovery of populations ofspecialist birds may take longer particularly in forest

Fig 7 Weighted-mean effect sizes and 95 bias-corrected confidence intervals of fire on soil properties from studies published between 1990and 2019 that were used in the review of fire effects across South American ecosystems to determine the role of climate and time since fire (A)Effect sizes of bulk density litter organic matter (OM) total nitrogen (Total N) available nitrogen (N) phosphorous (P) micronutrients pH andsalinity are shown (B) Effect sizes of early (-early) and late (-late) post-fire time of organic matter total nitrogen available nitrogen phosphorousand within climate type (-humid-warm -humid-cold -semiarid) of organic matter and total nitrogen Parameters with confidence intervals that donot overlap the vertical dotted line (Hedgersquos d = 0) are considered to have a significant positive or negative effect Sample sizes for each categoryare shown in parentheses The size of each black dot is proportional to its weight or contribution to the overall mean calculation Asterisk ()denotes a significant difference (Q-between [Qb] statistics) among categories ( = P lt 005)

Giorgis et al Fire Ecology (2021) 1711 Page 12 of 20

ecosystems (eg Barlow and Peres 2004 Fontaine andKennedy 2012 Mestre et al 2013 Albanesi et al 2014)although there are exceptions (eg Fontaine andKennedy 2012 Lee 2018 Morales et al 2020) such asthe group of birds known as cavity excavators (iewoodpeckers [Picidae]) that take advantage of both thesoftness of the wood of recently burned trees (ie snags)for making their cavities (Schepps et al 1999 Winklerand Christie 2002 Bond et al 2012 Lorenz et al 2015)and the higher amount of food (insect and other inverte-brates) found in that substrate right after the fire(Murphy and Lehnhausen 1998 Nappi et al 2003)Plant fitness Despite the negative effect of fire on tree

abundance we found no effect of fire on tree fitnessand a slight positive trend on vegetative fitness early postfire (Fig 6A B) Many tree species in the analyzed eco-systems are highly resilient to individual fires mainlythrough resprouting (Hoffmann and Solbrig 2003 Sou-chie et al 2017 Jaureguiberry et al 2020) therefore ifburned plots do not burn again trees might grow andrecover their biomass (Doherty et al 2017 Miller et al2019 Coop et al 2020) This response does not neces-sarily imply a direct adaptation to fire as the fire-adaptivetraits on which it relies could also have arisen in responseto other factors that have had a long history in some ofthe studied regions such as herbivory and seasonaldrought (Keeley et al 2011) which indirectly provide thevegetation with a high resilience to fire On the otherhand the negative tendency of the reproductive fitness ob-served early post fire is expected since trees may take along time to recover their reproductive size (Chapin IIIet al 2011) In contrast to trees shrubs did show an in-crease in vegetative and reproductive fitness early post firewhich differs significantly from that of late post fire(Fig 6C) The early vegetative response reflects the highresprouting capacity of many shrub species across thestudied regions especially in fire-prone regions (Gurvichet al 2005 Bravo et al 2014 Torres et al 2014 Duriganet al 2020 Jaureguiberry et al 2020) Regarding the repro-ductive response the pattern reported here agrees withprevious evidence suggesting that shrubs might reachminimum reproductive sizes soon after fire then slowdown their growth rate and reproduction as successionprogresses (Hoffmann and Moreira 2002 Hoffmann andSolbrig 2003 Galiacutendez et al 2009) In the case of herb-aceous plants surprisingly we did not find a general effectin fitness but a positive trend in reproductive fitness wasobserved mainly early post fire for grasses (Fig 6B C)The lack of a consistent general response pattern in forbscould be explained by the high heterogeneity of functionalgroups included within this growth form (eg annual per-ennial climber fern) that can have different responses tofire (Keeley et al 1981 Bates et al 2014 Heydari et al2016 Arcamone and Jaureguiberry 2018 Vidaller et al

2019) In the case of grasses the tendency of increasing re-productive fitness early post fire (Fig 6C) could be relatedto the positive firendashcolonization feedback generally de-scribed for this growth form (Bond et al 2003 Bond andKeeley 2005 Pausas and Keeley 2009 Pilon et al 2018)Previous studies had observed an increase in reproductivefitness of grasses after fire (Baruch and Bilbao 1999 Ara-uacutejo et al 2013 Pilon et al 2018 Vidaller et al 2019) Onthe contrary vegetative fitness showed a negative impactearly post fire However we cannot determine if the ob-served pattern (ie positive reproductive fitness versusnegative vegetative fitness) was due to intrinsic differencesbetween types of fitness or to differences in the type of cli-mate since all studies on reproductive fitness were lo-cated in humid-warm climate while vegetative fitnesswas mostly studied in semiarid and arid climatesSoil properties The responses of soil properties to fire in

South American ecosystems support the trend suggestedin previous global reviews (Certini 2005 Pellegrini et al2018) Fire significantly reduces the amount of two im-portant soil properties OM and nitrogen content (Pelle-grini et al 2018) On the one hand OM is directlyassociated with carbon concentration in the soil one ofthe most stable carbon reservoirs worldwide thus large-scale losses of OM due to fire could have important impli-cations for global climate change (Lal 2004) On the otherhand nitrogen is often the most limiting soil nutrient forplant growth and its loss directly affects ecosystems by re-ducing net primary productivity (Vitousek and Howarth1991 Pellegrini et al 2018) In addition our results of ahigher negative effect of fire on OM in semiarid cli-mates compared to humid-warm climate is supportedby Pellegrini et al (2018) who found that across sa-vannas and grasslands biannual fire simulation induceshigh carbon losses under drier climates However ithas been recently proposed that a single fire event maybe related to either an increase a decrease or nochange in soil carbon and nitrogen while frequent firesover time produce a consistent decrease of both ele-ments across ecosystems (Pellegrini et al 2020a b)This hypothesis poses a constraint to our analysessince most of the articles did not report fire history fortheir study sites Consequently the heterogeneous ef-fects of single fires on soil carbon and nitrogen may beconfused with those of frequent fires therefore limitingthe identification of clearer patternsFire produces an increase in bulk density as a result

of soil disaggregation due to the decrease or loss ofOM and to the infiltration of ashes into soil micro-pores (Boyer and Miller 1994 Mataix-Solera et al2011) Soil pH also increased in burned plots as aresult of organic acid denaturation due to soil heatingand the accumulation of potassium and sodiumhydroxides and magnesium and calcium carbonates

Giorgis et al Fire Ecology (2021) 1711 Page 13 of 20

(Knicker 2007) The increase in pH is also driven bythe reduction in OM (Certini 2005) Additionallyimmediate phosphorus enrichment after a fire event isthe result of organic pool conversion of phosphorusinto orthophosphate which declines at later post-firetimes as it is also suggested by Certini (2005)Climatendashfirendashvegetation feedbacks Our review sug-

gests that climatic conditions associated with higherrates of biomass accumulation might determine a fasterecosystem recovery in burned sites across South Amer-ica We found a faster recovery in invertebrate abun-dance and lower losses of soil organic matter underclimates with higher water availability and temperaturesthan in semiarid climates Beyond climatic conditions inhistorically fire-prone ecosystems where fire has pre-sumably been an important evolutionary factor (Bondand Keeley 2005) a greater recovery of the biota is ex-pected compared to non-fire-prone ecosystems Howeverour classification of climate types may not necessarily re-flect the fire history of the corresponding regions there-fore making a clear interpretation of the reported patternsdifficult Although semiarid climate sites could be clearlyconsidered as fire-prone systems (Bond 2005) warm-humid climate sites may also include fire-prone ecosys-tems because some of them have seasonal rainfall pat-terns which is a key characteristic in determining the fireproneness of an ecosystem (Nogueira et al 2017 Romanoand Ursino 2020) In fact following the classification ofOlson et al (2001) 40 of the studies from warm-humidareas corresponded to tropical and subtropical grasslandsavanna and shrublands biome (arguably fire prone)while the other 60 corresponded to tropical and sub-tropical moist broadleaf forest biome (arguably non-fireprone) Most of the former were located in the Cerradoregion where Hoffmann and Moreira (2002) found a fas-ter post-fire recovery of woody species compared toneighboring tropical forest areas Thus they concludedthat the combination of fire-prone vegetation presumablywith a long fire history and warm-humid climate mightfavor a fast ecosystem recovery after fire Therefore theobserved results of greater recovery in warm-humid eco-systems types is likely due to a combination of both cli-mate and fire history Likewise most of the studies fromcold-humid areas have a mediterranean-type climate (egnorthwest Patagonia in Argentina Keeley et al 2012)where rainy winters and hot and dry summers make theseregions fire prone Such regions have been historicallysubjected to periodic fires (both natural and anthropo-genic) and therefore many plant species are capable of re-covering after a fire event (Veblen et al 2003 2008Defosseacute et al 2015) Interestingly we found a significantpost-fire decrease in nitrogen content and soil OM undersemiarid climate but not under cold-humid climateThis response pattern may be due to the inherent

differences in ecosystem properties between climatetypes (eg decomposition and growth rates succes-sional patterns and fire regimen among othersMcLauchlan et al 2020) Fire regime might play animportant role as more frequent surface fire in semi-arid ecosystems could lead to a lower recovery of soilproperties compared to cold-humid ecosystems wherecrown fires predominate and fire frequency is lower thanin semiarid ecosystems therefore buffering fire effects onsoil propertiesThe lack of a standardized classification of fire-

prone and non-fire-prone regions as well as the lim-ited available information on the frequency intensityand severity of fires in the research studies includedin this review are certainly limitations that call forprecaution in the interpretation and extrapolation ofthe results reported here Overcoming such limita-tions remains a challenge in fire ecology (Harris et al2016 Kelly et al 2018) We believe that our studyrepresents a step forward in the synthesis of fire-related patterns at a large scale Our meta-analysis aswell as the discussed patterns provides relevant infor-mation toward understanding the feedback betweenclimate vegetation and fire on the South Americancontinent An ultimately desirable goal in the mediumor long term would be the development of a globalnetwork of fire field experimental surveys across sys-tems with different vegetation types and fire histories

Future perspectivesAcross recently published global fire reviews there isclearly a low representation of studies from SouthAmerica (Prichard et al 2017 Geary et al 2019)Given the relevance of this continent in the globaldynamics of fire it is highly necessary that morestudies on different aspects of fire ecology be carriedout throughout fire-prone regions of the continent(eg seasonal forests and humid and semiarid savannas)Furthermore our review allowed us to identify specific re-search gaps which can be added to the agenda of researchpriorities for future studies in fire ecology (McLauchlanet al 2020)Most of the analyzed studies failed to include fire

characteristics such as frequency intensity and severitywhich prevented us from obtaining a reliable pattern oftheir role as modulators of ecosystem and organismresponses to fire This is an unmet challenge in fireresearch as these fire characteristics may certainly influ-ence the trends reported in this review (Keeley 2009Fontaine and Kennedy 2012 Balch et al 2015 Silveiraet al 2016 Carbone et al 2019)Considering that woody ecosystems might take more

than 40 years to recover after fire (Cavallero et al 2015

Giorgis et al Fire Ecology (2021) 1711 Page 14 of 20

Doherty et al 2017) studies on fire responses that spana short recovery period fail to elucidate what might hap-pen in the meantime between fire occurrence and recov-ery Consequently our review poses an urgent call toincrease the studies assessing changes in fire effects overlonger post-fire times than those usually consideredFurther studies focusing on biodiversity are needed to

increase sensibility to detect changes across taxa groupsclimate types and post-fire timesAdditionally there is a gap of studies on vertebrate

responses to fire and some of the few available ones havemethodological limitations which prevented us from in-cluding them in our analysis (eg lack of replicates inmammals studies Griffiths and Brook 2014)Studies dealing with fire and exotic or invasive species

were quite scarce While invasion ecology has shown in-creasing attention in the last decades (eg Gurevitch et al2011 Pyšek et al 2020) it does not seem to be reflected infire studies in South America (but see Chaneton et al2004 Hoffmann et al 2004 Maziacutea et al 2010 Raffaeleet al 2016 Herrero et al 2016 Marcora et al 2018) Webelieve that this should be a priority topic for future re-search particularly considering the tight link betweenplant invasions and changes in fire regimes (Brooks et al2004 Mandle et al 2011 Harris et al 2016)

Supplementary InformationThe online version contains supplementary material available at httpsdoiorg101186s42408-021-00100-9

Additional file 1 Supplementary material regarding the choice ofstudies published between 1990 and 2019 that were used in the reviewof fire effects across South American ecosystems to determine the role ofclimate and time since fire Figure 11 PRISMA flow diagramrepresenting the flow of information through the decision process (iethe number of studies identified rejected and accepted) for the meta-analysis of fire effects across South American ecosystems to determinethe role of climate and time since fire Studies used were published be-tween 1990 and 2019 List 11 List of reviewed studies published be-tween 1990 and 2019 that were used in the meta-analysis of fire effectsacross South American ecosystems to determine the role of climate andtime since fire

Additional file 2 Distribution across the climatic temperaturendashprecipitation plane of study sites (n = 175) from the 160 articles reviewedfor the meta-analysis of fire effects across South American ecosystems todetermine the role of climate and time since fire Studies used were pub-lished between 1990 and 2019 The four climate types established for thecollected data are indicated by different colors

Additional file 3 Distribution across the climatic temperaturendashprecipitation plane of study sites (n = 175) from the 160 articles reviewedfor the meta-analysis of fire effects across South American ecosystems todetermine the role of climate and time since fire Studies used were pub-lished between 1990 and 2019 The four climate types established for thecollected data are indicated by different colors Table 31 Synopsis ofmodels used in the different hierarchical mixed effects meta-analyses offire effects across South American ecosystems taken from studies pub-lished between 1990 and 2019 to determine the role of climate and timesince fire expressed in R language syntax using the metafor package(Viechtbauer 2010) yi = Hedgesrsquo d vi = variance of Hedgesrsquo d ID = effectidentity Table 32 Summary of the fire effects on biodiversity abun-dance fitness and soil properties showing tests of moderator variables

and heterogeneities from the meta-analyses of fire effects across SouthAmerican ecosystems to determine the role of climate and time sincefire Studies used were published between 1990 and 2019 Significant in-fluence of moderators is indicated in bold when P le 005 and the 95confidence interval (CI) does not include zero Non-significant effects (P gt005) are indicated with ldquonsrdquo Note that Q-total (QT) statistics are reportedfor overall effect and Q-between (Qb) statistics are reported for categor-ical moderator variables

Additional file 4 Supplementary information 41 Publication biasfor the meta-analysis of fire effects across South American ecosystems todetermine the role of climate and time since fire Studies used were pub-lished between 1990 and 2019 Table 41 Sample size (number of obser-vations) and results of publication bias and hierarchical models testingfor each response variable in our meta-analysis of fire effects across SouthAmerican ecosystems to determine the role of climate and time sincefire Studies used were published between 1990 and 2019 Publicationbias was tested using fail-safe number (ie Rosenberg method) Kendallrsquosrank correlation test (z) and ldquotrim and fillrdquo models

AcknowledgementsMAG and HvW were partially supported by the National Scientific andTechnological Research Council (CONICET) fellow program MAG SRZ LC RAAF PAT EK FB DEG PV and PJ are researchers of CONICET Argentina MAGLC PAT EK FB and DEG are professors at the National University of Cordobaand PV is professor at the Universidad Nacional de Cuyo

Authorsrsquo contributionsMAG SRZ LC RA HZ HvW and PJ designed the research and analysesMAG SRZ LC RA and PJ conducted literature search with the contributionof all the authors MAG LC and PJ wrote the article with contributions fromRA AF PT and EK All authors read and approved the final manuscript

FundingThe National Council for Scientific and Technological Research (CONICET) ofArgentina indirectly supported this manuscript through exchange grants forMAG and HvW However no direct funding supported this project

Availability of data and materialsNo unique data were created in the writing of this manuscript so noadditional data are available

Declarations

Ethics approval and consent to participateNot applicable

Consent for publicationConsent for publication not applicable as we use previously published

Competing interestsThe authors declare that they have no competing interests

Author details1Instituto Multidisciplinario de Biologia Vegetal (CONICET-UNC) VelezSarsfield 1611 CC 495 5000 Cordoba Argentina 2Facultad de CienciasExactas Fisicas y Naturales Universidad Nacional de Cordoba Velez Sarsfield299 CC 495 5000 Cordoba Argentina 3Faculty of Sustainability LeuphanaUniversity Luneburg Institute for Ethics and Transdisciplinary SustainabilityResearch Universitatsallee 1 21335 Luneburg Germany 4Faculty ofSustainability Leuphana University Luneburg Institute of EcologyUniversitatsallee 1 21335 Luneburg Germany 5Instituto de Diversidad yEcologia Animal (IDEA) CONICET-UNC and Facultad de Ciencias ExactasFisicas y Naturales Universidad Nacional de Cordoba Velez Sarsfield 299 CP5000 Cordoba Argentina 6Facultad de Ciencias Agrarias UniversidadNacional de Cuyo Brown sin nuacutemero Chacras de Coria 5505 MendozaArgentina 7Facultad de Ciencias Agrarias Universidad Nacional de CuyoM5528AHB Chacras de Coria Mendoza Argentina

Giorgis et al Fire Ecology (2021) 1711 Page 15 of 20

ecosystems (eg Barlow and Peres 2004 Fontaine andKennedy 2012 Mestre et al 2013 Albanesi et al 2014)although there are exceptions (eg Fontaine andKennedy 2012 Lee 2018 Morales et al 2020) such asthe group of birds known as cavity excavators (iewoodpeckers [Picidae]) that take advantage of both thesoftness of the wood of recently burned trees (ie snags)for making their cavities (Schepps et al 1999 Winklerand Christie 2002 Bond et al 2012 Lorenz et al 2015)and the higher amount of food (insect and other inverte-brates) found in that substrate right after the fire(Murphy and Lehnhausen 1998 Nappi et al 2003)Plant fitness Despite the negative effect of fire on tree

abundance we found no effect of fire on tree fitnessand a slight positive trend on vegetative fitness early postfire (Fig 6A B) Many tree species in the analyzed eco-systems are highly resilient to individual fires mainlythrough resprouting (Hoffmann and Solbrig 2003 Sou-chie et al 2017 Jaureguiberry et al 2020) therefore ifburned plots do not burn again trees might grow andrecover their biomass (Doherty et al 2017 Miller et al2019 Coop et al 2020) This response does not neces-sarily imply a direct adaptation to fire as the fire-adaptivetraits on which it relies could also have arisen in responseto other factors that have had a long history in some ofthe studied regions such as herbivory and seasonaldrought (Keeley et al 2011) which indirectly provide thevegetation with a high resilience to fire On the otherhand the negative tendency of the reproductive fitness ob-served early post fire is expected since trees may take along time to recover their reproductive size (Chapin IIIet al 2011) In contrast to trees shrubs did show an in-crease in vegetative and reproductive fitness early post firewhich differs significantly from that of late post fire(Fig 6C) The early vegetative response reflects the highresprouting capacity of many shrub species across thestudied regions especially in fire-prone regions (Gurvichet al 2005 Bravo et al 2014 Torres et al 2014 Duriganet al 2020 Jaureguiberry et al 2020) Regarding the repro-ductive response the pattern reported here agrees withprevious evidence suggesting that shrubs might reachminimum reproductive sizes soon after fire then slowdown their growth rate and reproduction as successionprogresses (Hoffmann and Moreira 2002 Hoffmann andSolbrig 2003 Galiacutendez et al 2009) In the case of herb-aceous plants surprisingly we did not find a general effectin fitness but a positive trend in reproductive fitness wasobserved mainly early post fire for grasses (Fig 6B C)The lack of a consistent general response pattern in forbscould be explained by the high heterogeneity of functionalgroups included within this growth form (eg annual per-ennial climber fern) that can have different responses tofire (Keeley et al 1981 Bates et al 2014 Heydari et al2016 Arcamone and Jaureguiberry 2018 Vidaller et al

2019) In the case of grasses the tendency of increasing re-productive fitness early post fire (Fig 6C) could be relatedto the positive firendashcolonization feedback generally de-scribed for this growth form (Bond et al 2003 Bond andKeeley 2005 Pausas and Keeley 2009 Pilon et al 2018)Previous studies had observed an increase in reproductivefitness of grasses after fire (Baruch and Bilbao 1999 Ara-uacutejo et al 2013 Pilon et al 2018 Vidaller et al 2019) Onthe contrary vegetative fitness showed a negative impactearly post fire However we cannot determine if the ob-served pattern (ie positive reproductive fitness versusnegative vegetative fitness) was due to intrinsic differencesbetween types of fitness or to differences in the type of cli-mate since all studies on reproductive fitness were lo-cated in humid-warm climate while vegetative fitnesswas mostly studied in semiarid and arid climatesSoil properties The responses of soil properties to fire in

South American ecosystems support the trend suggestedin previous global reviews (Certini 2005 Pellegrini et al2018) Fire significantly reduces the amount of two im-portant soil properties OM and nitrogen content (Pelle-grini et al 2018) On the one hand OM is directlyassociated with carbon concentration in the soil one ofthe most stable carbon reservoirs worldwide thus large-scale losses of OM due to fire could have important impli-cations for global climate change (Lal 2004) On the otherhand nitrogen is often the most limiting soil nutrient forplant growth and its loss directly affects ecosystems by re-ducing net primary productivity (Vitousek and Howarth1991 Pellegrini et al 2018) In addition our results of ahigher negative effect of fire on OM in semiarid cli-mates compared to humid-warm climate is supportedby Pellegrini et al (2018) who found that across sa-vannas and grasslands biannual fire simulation induceshigh carbon losses under drier climates However ithas been recently proposed that a single fire event maybe related to either an increase a decrease or nochange in soil carbon and nitrogen while frequent firesover time produce a consistent decrease of both ele-ments across ecosystems (Pellegrini et al 2020a b)This hypothesis poses a constraint to our analysessince most of the articles did not report fire history fortheir study sites Consequently the heterogeneous ef-fects of single fires on soil carbon and nitrogen may beconfused with those of frequent fires therefore limitingthe identification of clearer patternsFire produces an increase in bulk density as a result

of soil disaggregation due to the decrease or loss ofOM and to the infiltration of ashes into soil micro-pores (Boyer and Miller 1994 Mataix-Solera et al2011) Soil pH also increased in burned plots as aresult of organic acid denaturation due to soil heatingand the accumulation of potassium and sodiumhydroxides and magnesium and calcium carbonates

Giorgis et al Fire Ecology (2021) 1711 Page 13 of 20

(Knicker 2007) The increase in pH is also driven bythe reduction in OM (Certini 2005) Additionallyimmediate phosphorus enrichment after a fire event isthe result of organic pool conversion of phosphorusinto orthophosphate which declines at later post-firetimes as it is also suggested by Certini (2005)Climatendashfirendashvegetation feedbacks Our review sug-

gests that climatic conditions associated with higherrates of biomass accumulation might determine a fasterecosystem recovery in burned sites across South Amer-ica We found a faster recovery in invertebrate abun-dance and lower losses of soil organic matter underclimates with higher water availability and temperaturesthan in semiarid climates Beyond climatic conditions inhistorically fire-prone ecosystems where fire has pre-sumably been an important evolutionary factor (Bondand Keeley 2005) a greater recovery of the biota is ex-pected compared to non-fire-prone ecosystems Howeverour classification of climate types may not necessarily re-flect the fire history of the corresponding regions there-fore making a clear interpretation of the reported patternsdifficult Although semiarid climate sites could be clearlyconsidered as fire-prone systems (Bond 2005) warm-humid climate sites may also include fire-prone ecosys-tems because some of them have seasonal rainfall pat-terns which is a key characteristic in determining the fireproneness of an ecosystem (Nogueira et al 2017 Romanoand Ursino 2020) In fact following the classification ofOlson et al (2001) 40 of the studies from warm-humidareas corresponded to tropical and subtropical grasslandsavanna and shrublands biome (arguably fire prone)while the other 60 corresponded to tropical and sub-tropical moist broadleaf forest biome (arguably non-fireprone) Most of the former were located in the Cerradoregion where Hoffmann and Moreira (2002) found a fas-ter post-fire recovery of woody species compared toneighboring tropical forest areas Thus they concludedthat the combination of fire-prone vegetation presumablywith a long fire history and warm-humid climate mightfavor a fast ecosystem recovery after fire Therefore theobserved results of greater recovery in warm-humid eco-systems types is likely due to a combination of both cli-mate and fire history Likewise most of the studies fromcold-humid areas have a mediterranean-type climate (egnorthwest Patagonia in Argentina Keeley et al 2012)where rainy winters and hot and dry summers make theseregions fire prone Such regions have been historicallysubjected to periodic fires (both natural and anthropo-genic) and therefore many plant species are capable of re-covering after a fire event (Veblen et al 2003 2008Defosseacute et al 2015) Interestingly we found a significantpost-fire decrease in nitrogen content and soil OM undersemiarid climate but not under cold-humid climateThis response pattern may be due to the inherent

differences in ecosystem properties between climatetypes (eg decomposition and growth rates succes-sional patterns and fire regimen among othersMcLauchlan et al 2020) Fire regime might play animportant role as more frequent surface fire in semi-arid ecosystems could lead to a lower recovery of soilproperties compared to cold-humid ecosystems wherecrown fires predominate and fire frequency is lower thanin semiarid ecosystems therefore buffering fire effects onsoil propertiesThe lack of a standardized classification of fire-

prone and non-fire-prone regions as well as the lim-ited available information on the frequency intensityand severity of fires in the research studies includedin this review are certainly limitations that call forprecaution in the interpretation and extrapolation ofthe results reported here Overcoming such limita-tions remains a challenge in fire ecology (Harris et al2016 Kelly et al 2018) We believe that our studyrepresents a step forward in the synthesis of fire-related patterns at a large scale Our meta-analysis aswell as the discussed patterns provides relevant infor-mation toward understanding the feedback betweenclimate vegetation and fire on the South Americancontinent An ultimately desirable goal in the mediumor long term would be the development of a globalnetwork of fire field experimental surveys across sys-tems with different vegetation types and fire histories

Future perspectivesAcross recently published global fire reviews there isclearly a low representation of studies from SouthAmerica (Prichard et al 2017 Geary et al 2019)Given the relevance of this continent in the globaldynamics of fire it is highly necessary that morestudies on different aspects of fire ecology be carriedout throughout fire-prone regions of the continent(eg seasonal forests and humid and semiarid savannas)Furthermore our review allowed us to identify specific re-search gaps which can be added to the agenda of researchpriorities for future studies in fire ecology (McLauchlanet al 2020)Most of the analyzed studies failed to include fire

characteristics such as frequency intensity and severitywhich prevented us from obtaining a reliable pattern oftheir role as modulators of ecosystem and organismresponses to fire This is an unmet challenge in fireresearch as these fire characteristics may certainly influ-ence the trends reported in this review (Keeley 2009Fontaine and Kennedy 2012 Balch et al 2015 Silveiraet al 2016 Carbone et al 2019)Considering that woody ecosystems might take more

than 40 years to recover after fire (Cavallero et al 2015

Giorgis et al Fire Ecology (2021) 1711 Page 14 of 20

Doherty et al 2017) studies on fire responses that spana short recovery period fail to elucidate what might hap-pen in the meantime between fire occurrence and recov-ery Consequently our review poses an urgent call toincrease the studies assessing changes in fire effects overlonger post-fire times than those usually consideredFurther studies focusing on biodiversity are needed to

increase sensibility to detect changes across taxa groupsclimate types and post-fire timesAdditionally there is a gap of studies on vertebrate

responses to fire and some of the few available ones havemethodological limitations which prevented us from in-cluding them in our analysis (eg lack of replicates inmammals studies Griffiths and Brook 2014)Studies dealing with fire and exotic or invasive species

were quite scarce While invasion ecology has shown in-creasing attention in the last decades (eg Gurevitch et al2011 Pyšek et al 2020) it does not seem to be reflected infire studies in South America (but see Chaneton et al2004 Hoffmann et al 2004 Maziacutea et al 2010 Raffaeleet al 2016 Herrero et al 2016 Marcora et al 2018) Webelieve that this should be a priority topic for future re-search particularly considering the tight link betweenplant invasions and changes in fire regimes (Brooks et al2004 Mandle et al 2011 Harris et al 2016)

Supplementary InformationThe online version contains supplementary material available at httpsdoiorg101186s42408-021-00100-9

Additional file 1 Supplementary material regarding the choice ofstudies published between 1990 and 2019 that were used in the reviewof fire effects across South American ecosystems to determine the role ofclimate and time since fire Figure 11 PRISMA flow diagramrepresenting the flow of information through the decision process (iethe number of studies identified rejected and accepted) for the meta-analysis of fire effects across South American ecosystems to determinethe role of climate and time since fire Studies used were published be-tween 1990 and 2019 List 11 List of reviewed studies published be-tween 1990 and 2019 that were used in the meta-analysis of fire effectsacross South American ecosystems to determine the role of climate andtime since fire

Additional file 2 Distribution across the climatic temperaturendashprecipitation plane of study sites (n = 175) from the 160 articles reviewedfor the meta-analysis of fire effects across South American ecosystems todetermine the role of climate and time since fire Studies used were pub-lished between 1990 and 2019 The four climate types established for thecollected data are indicated by different colors

Additional file 3 Distribution across the climatic temperaturendashprecipitation plane of study sites (n = 175) from the 160 articles reviewedfor the meta-analysis of fire effects across South American ecosystems todetermine the role of climate and time since fire Studies used were pub-lished between 1990 and 2019 The four climate types established for thecollected data are indicated by different colors Table 31 Synopsis ofmodels used in the different hierarchical mixed effects meta-analyses offire effects across South American ecosystems taken from studies pub-lished between 1990 and 2019 to determine the role of climate and timesince fire expressed in R language syntax using the metafor package(Viechtbauer 2010) yi = Hedgesrsquo d vi = variance of Hedgesrsquo d ID = effectidentity Table 32 Summary of the fire effects on biodiversity abun-dance fitness and soil properties showing tests of moderator variables

and heterogeneities from the meta-analyses of fire effects across SouthAmerican ecosystems to determine the role of climate and time sincefire Studies used were published between 1990 and 2019 Significant in-fluence of moderators is indicated in bold when P le 005 and the 95confidence interval (CI) does not include zero Non-significant effects (P gt005) are indicated with ldquonsrdquo Note that Q-total (QT) statistics are reportedfor overall effect and Q-between (Qb) statistics are reported for categor-ical moderator variables

Additional file 4 Supplementary information 41 Publication biasfor the meta-analysis of fire effects across South American ecosystems todetermine the role of climate and time since fire Studies used were pub-lished between 1990 and 2019 Table 41 Sample size (number of obser-vations) and results of publication bias and hierarchical models testingfor each response variable in our meta-analysis of fire effects across SouthAmerican ecosystems to determine the role of climate and time sincefire Studies used were published between 1990 and 2019 Publicationbias was tested using fail-safe number (ie Rosenberg method) Kendallrsquosrank correlation test (z) and ldquotrim and fillrdquo models

AcknowledgementsMAG and HvW were partially supported by the National Scientific andTechnological Research Council (CONICET) fellow program MAG SRZ LC RAAF PAT EK FB DEG PV and PJ are researchers of CONICET Argentina MAGLC PAT EK FB and DEG are professors at the National University of Cordobaand PV is professor at the Universidad Nacional de Cuyo

Authorsrsquo contributionsMAG SRZ LC RA HZ HvW and PJ designed the research and analysesMAG SRZ LC RA and PJ conducted literature search with the contributionof all the authors MAG LC and PJ wrote the article with contributions fromRA AF PT and EK All authors read and approved the final manuscript

FundingThe National Council for Scientific and Technological Research (CONICET) ofArgentina indirectly supported this manuscript through exchange grants forMAG and HvW However no direct funding supported this project

Availability of data and materialsNo unique data were created in the writing of this manuscript so noadditional data are available

Declarations

Ethics approval and consent to participateNot applicable

Consent for publicationConsent for publication not applicable as we use previously published

Competing interestsThe authors declare that they have no competing interests

Author details1Instituto Multidisciplinario de Biologia Vegetal (CONICET-UNC) VelezSarsfield 1611 CC 495 5000 Cordoba Argentina 2Facultad de CienciasExactas Fisicas y Naturales Universidad Nacional de Cordoba Velez Sarsfield299 CC 495 5000 Cordoba Argentina 3Faculty of Sustainability LeuphanaUniversity Luneburg Institute for Ethics and Transdisciplinary SustainabilityResearch Universitatsallee 1 21335 Luneburg Germany 4Faculty ofSustainability Leuphana University Luneburg Institute of EcologyUniversitatsallee 1 21335 Luneburg Germany 5Instituto de Diversidad yEcologia Animal (IDEA) CONICET-UNC and Facultad de Ciencias ExactasFisicas y Naturales Universidad Nacional de Cordoba Velez Sarsfield 299 CP5000 Cordoba Argentina 6Facultad de Ciencias Agrarias UniversidadNacional de Cuyo Brown sin nuacutemero Chacras de Coria 5505 MendozaArgentina 7Facultad de Ciencias Agrarias Universidad Nacional de CuyoM5528AHB Chacras de Coria Mendoza Argentina

Giorgis et al Fire Ecology (2021) 1711 Page 15 of 20

(Knicker 2007) The increase in pH is also driven bythe reduction in OM (Certini 2005) Additionallyimmediate phosphorus enrichment after a fire event isthe result of organic pool conversion of phosphorusinto orthophosphate which declines at later post-firetimes as it is also suggested by Certini (2005)Climatendashfirendashvegetation feedbacks Our review sug-

gests that climatic conditions associated with higherrates of biomass accumulation might determine a fasterecosystem recovery in burned sites across South Amer-ica We found a faster recovery in invertebrate abun-dance and lower losses of soil organic matter underclimates with higher water availability and temperaturesthan in semiarid climates Beyond climatic conditions inhistorically fire-prone ecosystems where fire has pre-sumably been an important evolutionary factor (Bondand Keeley 2005) a greater recovery of the biota is ex-pected compared to non-fire-prone ecosystems Howeverour classification of climate types may not necessarily re-flect the fire history of the corresponding regions there-fore making a clear interpretation of the reported patternsdifficult Although semiarid climate sites could be clearlyconsidered as fire-prone systems (Bond 2005) warm-humid climate sites may also include fire-prone ecosys-tems because some of them have seasonal rainfall pat-terns which is a key characteristic in determining the fireproneness of an ecosystem (Nogueira et al 2017 Romanoand Ursino 2020) In fact following the classification ofOlson et al (2001) 40 of the studies from warm-humidareas corresponded to tropical and subtropical grasslandsavanna and shrublands biome (arguably fire prone)while the other 60 corresponded to tropical and sub-tropical moist broadleaf forest biome (arguably non-fireprone) Most of the former were located in the Cerradoregion where Hoffmann and Moreira (2002) found a fas-ter post-fire recovery of woody species compared toneighboring tropical forest areas Thus they concludedthat the combination of fire-prone vegetation presumablywith a long fire history and warm-humid climate mightfavor a fast ecosystem recovery after fire Therefore theobserved results of greater recovery in warm-humid eco-systems types is likely due to a combination of both cli-mate and fire history Likewise most of the studies fromcold-humid areas have a mediterranean-type climate (egnorthwest Patagonia in Argentina Keeley et al 2012)where rainy winters and hot and dry summers make theseregions fire prone Such regions have been historicallysubjected to periodic fires (both natural and anthropo-genic) and therefore many plant species are capable of re-covering after a fire event (Veblen et al 2003 2008Defosseacute et al 2015) Interestingly we found a significantpost-fire decrease in nitrogen content and soil OM undersemiarid climate but not under cold-humid climateThis response pattern may be due to the inherent

differences in ecosystem properties between climatetypes (eg decomposition and growth rates succes-sional patterns and fire regimen among othersMcLauchlan et al 2020) Fire regime might play animportant role as more frequent surface fire in semi-arid ecosystems could lead to a lower recovery of soilproperties compared to cold-humid ecosystems wherecrown fires predominate and fire frequency is lower thanin semiarid ecosystems therefore buffering fire effects onsoil propertiesThe lack of a standardized classification of fire-

prone and non-fire-prone regions as well as the lim-ited available information on the frequency intensityand severity of fires in the research studies includedin this review are certainly limitations that call forprecaution in the interpretation and extrapolation ofthe results reported here Overcoming such limita-tions remains a challenge in fire ecology (Harris et al2016 Kelly et al 2018) We believe that our studyrepresents a step forward in the synthesis of fire-related patterns at a large scale Our meta-analysis aswell as the discussed patterns provides relevant infor-mation toward understanding the feedback betweenclimate vegetation and fire on the South Americancontinent An ultimately desirable goal in the mediumor long term would be the development of a globalnetwork of fire field experimental surveys across sys-tems with different vegetation types and fire histories

Future perspectivesAcross recently published global fire reviews there isclearly a low representation of studies from SouthAmerica (Prichard et al 2017 Geary et al 2019)Given the relevance of this continent in the globaldynamics of fire it is highly necessary that morestudies on different aspects of fire ecology be carriedout throughout fire-prone regions of the continent(eg seasonal forests and humid and semiarid savannas)Furthermore our review allowed us to identify specific re-search gaps which can be added to the agenda of researchpriorities for future studies in fire ecology (McLauchlanet al 2020)Most of the analyzed studies failed to include fire

characteristics such as frequency intensity and severitywhich prevented us from obtaining a reliable pattern oftheir role as modulators of ecosystem and organismresponses to fire This is an unmet challenge in fireresearch as these fire characteristics may certainly influ-ence the trends reported in this review (Keeley 2009Fontaine and Kennedy 2012 Balch et al 2015 Silveiraet al 2016 Carbone et al 2019)Considering that woody ecosystems might take more

than 40 years to recover after fire (Cavallero et al 2015

Giorgis et al Fire Ecology (2021) 1711 Page 14 of 20

Doherty et al 2017) studies on fire responses that spana short recovery period fail to elucidate what might hap-pen in the meantime between fire occurrence and recov-ery Consequently our review poses an urgent call toincrease the studies assessing changes in fire effects overlonger post-fire times than those usually consideredFurther studies focusing on biodiversity are needed to

increase sensibility to detect changes across taxa groupsclimate types and post-fire timesAdditionally there is a gap of studies on vertebrate

responses to fire and some of the few available ones havemethodological limitations which prevented us from in-cluding them in our analysis (eg lack of replicates inmammals studies Griffiths and Brook 2014)Studies dealing with fire and exotic or invasive species

were quite scarce While invasion ecology has shown in-creasing attention in the last decades (eg Gurevitch et al2011 Pyšek et al 2020) it does not seem to be reflected infire studies in South America (but see Chaneton et al2004 Hoffmann et al 2004 Maziacutea et al 2010 Raffaeleet al 2016 Herrero et al 2016 Marcora et al 2018) Webelieve that this should be a priority topic for future re-search particularly considering the tight link betweenplant invasions and changes in fire regimes (Brooks et al2004 Mandle et al 2011 Harris et al 2016)

Supplementary InformationThe online version contains supplementary material available at httpsdoiorg101186s42408-021-00100-9

Additional file 1 Supplementary material regarding the choice ofstudies published between 1990 and 2019 that were used in the reviewof fire effects across South American ecosystems to determine the role ofclimate and time since fire Figure 11 PRISMA flow diagramrepresenting the flow of information through the decision process (iethe number of studies identified rejected and accepted) for the meta-analysis of fire effects across South American ecosystems to determinethe role of climate and time since fire Studies used were published be-tween 1990 and 2019 List 11 List of reviewed studies published be-tween 1990 and 2019 that were used in the meta-analysis of fire effectsacross South American ecosystems to determine the role of climate andtime since fire

Additional file 2 Distribution across the climatic temperaturendashprecipitation plane of study sites (n = 175) from the 160 articles reviewedfor the meta-analysis of fire effects across South American ecosystems todetermine the role of climate and time since fire Studies used were pub-lished between 1990 and 2019 The four climate types established for thecollected data are indicated by different colors

Additional file 3 Distribution across the climatic temperaturendashprecipitation plane of study sites (n = 175) from the 160 articles reviewedfor the meta-analysis of fire effects across South American ecosystems todetermine the role of climate and time since fire Studies used were pub-lished between 1990 and 2019 The four climate types established for thecollected data are indicated by different colors Table 31 Synopsis ofmodels used in the different hierarchical mixed effects meta-analyses offire effects across South American ecosystems taken from studies pub-lished between 1990 and 2019 to determine the role of climate and timesince fire expressed in R language syntax using the metafor package(Viechtbauer 2010) yi = Hedgesrsquo d vi = variance of Hedgesrsquo d ID = effectidentity Table 32 Summary of the fire effects on biodiversity abun-dance fitness and soil properties showing tests of moderator variables

and heterogeneities from the meta-analyses of fire effects across SouthAmerican ecosystems to determine the role of climate and time sincefire Studies used were published between 1990 and 2019 Significant in-fluence of moderators is indicated in bold when P le 005 and the 95confidence interval (CI) does not include zero Non-significant effects (P gt005) are indicated with ldquonsrdquo Note that Q-total (QT) statistics are reportedfor overall effect and Q-between (Qb) statistics are reported for categor-ical moderator variables

Additional file 4 Supplementary information 41 Publication biasfor the meta-analysis of fire effects across South American ecosystems todetermine the role of climate and time since fire Studies used were pub-lished between 1990 and 2019 Table 41 Sample size (number of obser-vations) and results of publication bias and hierarchical models testingfor each response variable in our meta-analysis of fire effects across SouthAmerican ecosystems to determine the role of climate and time sincefire Studies used were published between 1990 and 2019 Publicationbias was tested using fail-safe number (ie Rosenberg method) Kendallrsquosrank correlation test (z) and ldquotrim and fillrdquo models

AcknowledgementsMAG and HvW were partially supported by the National Scientific andTechnological Research Council (CONICET) fellow program MAG SRZ LC RAAF PAT EK FB DEG PV and PJ are researchers of CONICET Argentina MAGLC PAT EK FB and DEG are professors at the National University of Cordobaand PV is professor at the Universidad Nacional de Cuyo

Authorsrsquo contributionsMAG SRZ LC RA HZ HvW and PJ designed the research and analysesMAG SRZ LC RA and PJ conducted literature search with the contributionof all the authors MAG LC and PJ wrote the article with contributions fromRA AF PT and EK All authors read and approved the final manuscript

FundingThe National Council for Scientific and Technological Research (CONICET) ofArgentina indirectly supported this manuscript through exchange grants forMAG and HvW However no direct funding supported this project

Availability of data and materialsNo unique data were created in the writing of this manuscript so noadditional data are available

Declarations

Ethics approval and consent to participateNot applicable

Consent for publicationConsent for publication not applicable as we use previously published

Competing interestsThe authors declare that they have no competing interests

Author details1Instituto Multidisciplinario de Biologia Vegetal (CONICET-UNC) VelezSarsfield 1611 CC 495 5000 Cordoba Argentina 2Facultad de CienciasExactas Fisicas y Naturales Universidad Nacional de Cordoba Velez Sarsfield299 CC 495 5000 Cordoba Argentina 3Faculty of Sustainability LeuphanaUniversity Luneburg Institute for Ethics and Transdisciplinary SustainabilityResearch Universitatsallee 1 21335 Luneburg Germany 4Faculty ofSustainability Leuphana University Luneburg Institute of EcologyUniversitatsallee 1 21335 Luneburg Germany 5Instituto de Diversidad yEcologia Animal (IDEA) CONICET-UNC and Facultad de Ciencias ExactasFisicas y Naturales Universidad Nacional de Cordoba Velez Sarsfield 299 CP5000 Cordoba Argentina 6Facultad de Ciencias Agrarias UniversidadNacional de Cuyo Brown sin nuacutemero Chacras de Coria 5505 MendozaArgentina 7Facultad de Ciencias Agrarias Universidad Nacional de CuyoM5528AHB Chacras de Coria Mendoza Argentina

Giorgis et al Fire Ecology (2021) 1711 Page 15 of 20

Doherty et al 2017) studies on fire responses that spana short recovery period fail to elucidate what might hap-pen in the meantime between fire occurrence and recov-ery Consequently our review poses an urgent call toincrease the studies assessing changes in fire effects overlonger post-fire times than those usually consideredFurther studies focusing on biodiversity are needed to

increase sensibility to detect changes across taxa groupsclimate types and post-fire timesAdditionally there is a gap of studies on vertebrate

responses to fire and some of the few available ones havemethodological limitations which prevented us from in-cluding them in our analysis (eg lack of replicates inmammals studies Griffiths and Brook 2014)Studies dealing with fire and exotic or invasive species

were quite scarce While invasion ecology has shown in-creasing attention in the last decades (eg Gurevitch et al2011 Pyšek et al 2020) it does not seem to be reflected infire studies in South America (but see Chaneton et al2004 Hoffmann et al 2004 Maziacutea et al 2010 Raffaeleet al 2016 Herrero et al 2016 Marcora et al 2018) Webelieve that this should be a priority topic for future re-search particularly considering the tight link betweenplant invasions and changes in fire regimes (Brooks et al2004 Mandle et al 2011 Harris et al 2016)

Supplementary InformationThe online version contains supplementary material available at httpsdoiorg101186s42408-021-00100-9

Additional file 1 Supplementary material regarding the choice ofstudies published between 1990 and 2019 that were used in the reviewof fire effects across South American ecosystems to determine the role ofclimate and time since fire Figure 11 PRISMA flow diagramrepresenting the flow of information through the decision process (iethe number of studies identified rejected and accepted) for the meta-analysis of fire effects across South American ecosystems to determinethe role of climate and time since fire Studies used were published be-tween 1990 and 2019 List 11 List of reviewed studies published be-tween 1990 and 2019 that were used in the meta-analysis of fire effectsacross South American ecosystems to determine the role of climate andtime since fire

Additional file 2 Distribution across the climatic temperaturendashprecipitation plane of study sites (n = 175) from the 160 articles reviewedfor the meta-analysis of fire effects across South American ecosystems todetermine the role of climate and time since fire Studies used were pub-lished between 1990 and 2019 The four climate types established for thecollected data are indicated by different colors

Additional file 3 Distribution across the climatic temperaturendashprecipitation plane of study sites (n = 175) from the 160 articles reviewedfor the meta-analysis of fire effects across South American ecosystems todetermine the role of climate and time since fire Studies used were pub-lished between 1990 and 2019 The four climate types established for thecollected data are indicated by different colors Table 31 Synopsis ofmodels used in the different hierarchical mixed effects meta-analyses offire effects across South American ecosystems taken from studies pub-lished between 1990 and 2019 to determine the role of climate and timesince fire expressed in R language syntax using the metafor package(Viechtbauer 2010) yi = Hedgesrsquo d vi = variance of Hedgesrsquo d ID = effectidentity Table 32 Summary of the fire effects on biodiversity abun-dance fitness and soil properties showing tests of moderator variables

and heterogeneities from the meta-analyses of fire effects across SouthAmerican ecosystems to determine the role of climate and time sincefire Studies used were published between 1990 and 2019 Significant in-fluence of moderators is indicated in bold when P le 005 and the 95confidence interval (CI) does not include zero Non-significant effects (P gt005) are indicated with ldquonsrdquo Note that Q-total (QT) statistics are reportedfor overall effect and Q-between (Qb) statistics are reported for categor-ical moderator variables

Additional file 4 Supplementary information 41 Publication biasfor the meta-analysis of fire effects across South American ecosystems todetermine the role of climate and time since fire Studies used were pub-lished between 1990 and 2019 Table 41 Sample size (number of obser-vations) and results of publication bias and hierarchical models testingfor each response variable in our meta-analysis of fire effects across SouthAmerican ecosystems to determine the role of climate and time sincefire Studies used were published between 1990 and 2019 Publicationbias was tested using fail-safe number (ie Rosenberg method) Kendallrsquosrank correlation test (z) and ldquotrim and fillrdquo models

AcknowledgementsMAG and HvW were partially supported by the National Scientific andTechnological Research Council (CONICET) fellow program MAG SRZ LC RAAF PAT EK FB DEG PV and PJ are researchers of CONICET Argentina MAGLC PAT EK FB and DEG are professors at the National University of Cordobaand PV is professor at the Universidad Nacional de Cuyo

Authorsrsquo contributionsMAG SRZ LC RA HZ HvW and PJ designed the research and analysesMAG SRZ LC RA and PJ conducted literature search with the contributionof all the authors MAG LC and PJ wrote the article with contributions fromRA AF PT and EK All authors read and approved the final manuscript

FundingThe National Council for Scientific and Technological Research (CONICET) ofArgentina indirectly supported this manuscript through exchange grants forMAG and HvW However no direct funding supported this project

Availability of data and materialsNo unique data were created in the writing of this manuscript so noadditional data are available

Declarations

Ethics approval and consent to participateNot applicable

Consent for publicationConsent for publication not applicable as we use previously published

Competing interestsThe authors declare that they have no competing interests

Author details1Instituto Multidisciplinario de Biologia Vegetal (CONICET-UNC) VelezSarsfield 1611 CC 495 5000 Cordoba Argentina 2Facultad de CienciasExactas Fisicas y Naturales Universidad Nacional de Cordoba Velez Sarsfield299 CC 495 5000 Cordoba Argentina 3Faculty of Sustainability LeuphanaUniversity Luneburg Institute for Ethics and Transdisciplinary SustainabilityResearch Universitatsallee 1 21335 Luneburg Germany 4Faculty ofSustainability Leuphana University Luneburg Institute of EcologyUniversitatsallee 1 21335 Luneburg Germany 5Instituto de Diversidad yEcologia Animal (IDEA) CONICET-UNC and Facultad de Ciencias ExactasFisicas y Naturales Universidad Nacional de Cordoba Velez Sarsfield 299 CP5000 Cordoba Argentina 6Facultad de Ciencias Agrarias UniversidadNacional de Cuyo Brown sin nuacutemero Chacras de Coria 5505 MendozaArgentina 7Facultad de Ciencias Agrarias Universidad Nacional de CuyoM5528AHB Chacras de Coria Mendoza Argentina

Giorgis et al Fire Ecology (2021) 1711 Page 15 of 20

Received 28 October 2020 Accepted 16 March 2021

ReferencesAlbanesi S S Dardanelli and LM Bellis 2014 Effects of fire disturbance on

bird communities and species of mountain Serrano forest in centralArgentina Journal of Forest Research 19 105ndash114 httpsdoiorg101007s10310-012-0388-4

Andela N DC Morton L Giglio Y Chen GR van der Werf PS Kasibhatla RSDeFries GJ Collatz S Hantson S Kloster D Bachelet M Forrest G LasslopF Li S Mangeon JR Melton C Yue and JT Randerson 2017 A human-driven decline in global burned area Science 356 (6345) 1356ndash1362 httpsdoiorg101126scienceaal4108

Arauacutejo GM AF Amaral EM Bruna et al 2013 Fire drives the reproductiveresponses of herbaceous plants in a Neotropical swamp Plant Ecology 2141479ndash1484 httpsdoiorg101007s11258-013-0268-9

Arcamone JR and P Jaureguiberry 2018 Germination response of commonannual and perennial forbs to heat shock and smoke treatments in theChaco Serrano central Argentina Austral Ecology 43 567ndash577 httpsdoiorg101111aec12593

Archibald S CER Lehmann CM Belcher WJ Bond RA Bradstock A-L DaniauKG Dexter EJ Forrestel M Greve and T He 2018 Biological andgeophysical feedbacks with fire in the Earth system Environmental ResearchLetters 13 33003 httpsdoiorg1010881748-9326aa9ead

Archibald S CER Lehmann JL Goacutemez-Dans and RA Bradstock 2013Defining pyromes and global syndromes of fire regimes ProceedingsNational Academy of Science 110 (16) 6442ndash6447 httpsdoiorg101073pnas1211466110

Archibald S DP Roy BW van Wilgen and RJ Scholes 2009 What limits fireAn examination of drivers of burnt area in Southern Africa Global ChangeBiology 15 613ndash630 httpsdoiorg101111j1365-2486200801754x

Balch JK PM Brando DC Nepstad MT Coe D Silveacuterio TJ Massad EADavidson P Lefebvre C Oliveira-Santos W Rocha RTS Cury A Parsonsand KS Carvalho 2015 The susceptibility of southeastern Amazon forests tofire insights from a large-scale burn experiment Bioscience 65 893ndash905httpsdoiorg101093bioscibiv106

Barlow J and CA Peres 2004 Avifaunal responses to single and recurrentwildfires in Amazonian forests Ecological Applications 14 (5) 1358ndash1373httpsdoiorg10189003-5077

Barlow J and CA Peres 2008 Fire-mediated dieback and compositionalcascade in an Amazonian forest Philosophical Transactions of the RoyalSociety of London B Biological Sciences 363 1787ndash1794 httpsdoiorg101098rstb20070013

Barlow J CA Peres BO Lagan and T Haugaasen 2003 Large tree mortalityand the decline of forest biomass following Amazonian wildfires EcologyLetters 6 6ndash8 httpsdoiorg101046j1461-0248200300394x

Baruch Z and B Bilbao 1999 Effects of fire and defoliation on the life history ofnative and invader C4 grasses in a Neotropical savanna Oecologia 119 (4)510ndash520 httpsdoiorg101007s004420050814

Bates JD RN Sharp and KW Davies 2014 Sagebrush steppe recovery after firevaries by development phase of Juniperus occidentalis woodlandInternational Journal of Wildland Fire 23 117ndash130 httpsdoiorg101071WF12206

Bauer S and BJ Hoye 2014 Migratory animals couple biodiversity andecosystem functioning worldwide Science 344 (6179) 1242552 httpsdoiorg101126science1242552

Blackhall M E Raffaele J Paritsis F Tiribelli JM Morales T Kitzberger JH Gowdaand TT Veblen 2017 Effects of biological legacies and herbivory on fuels andflammability traits a long-term experimental study of alternative stable statesJournal of Ecology 105 1309ndash1322 httpsdoiorg1011111365-274512796

Blackhall M E Raffaele and TT Veblen 2015 Combined effects of fire and cattlein shrublands and forests of northwest Patagonia Ecologiacutea Austral 25 1ndash10httpsdoiorg1025260EA15251048

Bond ML RB Siegel and DL Craig eds 2012 A conservation strategy forthe black-backed woodpecker (Picoides arcticus) in California ndash Version 10Point Reyes Station The Institute for Bird Populations and CaliforniaPartners in Flight

Bond WJ 2005 Large parts of the world are brown or black a different view onthe lsquoGreen Worldrsquo hypothesis Journal of Vegetation Science 16 (3) 261ndash266

Bond W J amp B W Van Wilgen 2012 Fire and plants (Vol 14) Springer Science ampBusiness Media

Bond WJ and JE Keeley 2005 Fire as a global lsquoherbivorersquo the ecology andevolution of flammable ecosystems Trends in Ecological Evolution 20 387ndash394 httpsdoiorg101016jtree200504025

Bond WJ GF Midgley and FI Woodward 2003 What controls South Africanvegetation ndash climate or fire South African Journal of Botany 69 1ndash13 httpsdoiorg101016S0254-6299(15)30362-8

Bond WJ FI Woodward and GF Midgley 2005 The global distribution ofecosystems in a world without fire New Phytology 165 525ndash537 httpsdoiorg101111j1469-8137200401252x

Borenstein M H Cooper L Hedges and J Valentine 2009 Effect sizes forcontinuous data In The handbook of research syntheses and meta-analysis edH Cooper LV Hedges and JC Valentine 2nd ed 221ndash236 New YorkRussell Sage Foundation

Borenstein M LV Hedges JPT Higgins and HR Rothstein 2010 A basicintroduction to fixed-effect and random-effects models for meta-analysisResearch Synthesis Methods 1 (2) 97ndash111 httpsdoiorg101002jrsm12

Bowman DMJS JK Balch P Artaxo WJ Bond JM Carlson MA Cochrane CM DrsquoAntonio RS Defries JC Doyle SP Harrison FH Johnston JE KeeleyMA Krawchuk CA Kull JB Marston MA Moritz IC Prentice CI Roos ACScott TW Swetnam GR van der Werf and SJ Pyne 2009 Fire in the Earthsystem Science 324 (5926) 481ndash484 httpsdoiorg101126science1163886

Boyer WD and JH Miller 1994 Effect of burning and brush treatments on nutrientand soil physical properties in young longleaf pine stands Forest Ecology andManagement 70 (1-3) 311ndash318 httpsdoiorg1010160378-1127(94)90096-5

Bradstock RA 2010 A biogeographic model of fire regimes in Australia currentand future implications Global Ecology and Biogeography 19 145ndash158httpsdoiorg101111j1466-8238200900512x

Bravo S C Kunst M Leiva and R Ledesma 2014 Response of hardwood treeregeneration to surface fires western Chaco region Argentina Forest Ecologyand Management 326 36ndash45 httpsdoiorg101016jforeco201404009

Brooks M C DrsquoAntonio DM Richardson JB Grace JE Keeley JM DiTomaso RJ Hobbs M Pellant and D Pyke 2004 Effects of invasive alien plants on fireregimes Bioscience 54 (7) 677ndash688 httpsdoiorg1016410006-3568(2004)054[0677EOIAPO]20CO2

Bruno JF and BJ Cardinale 2008 Cascading effects of predator richness Frontiersin Ecology and the Environment 6 539ndash546 httpsdoiorg101890070136

Carbone LM and R Aguilar 2017 Fire frequency effects on soil and pollinatorswhat shapes sexual plant reproduction Plant Ecology 218 1283ndash1297httpsdoiorg101007s11258-017-0768-0

Carbone LM J Tavella JG Pausas and R Aguilar 2019 A global synthesis offire effects on pollinators Global Ecology and Biogeography 28 (10) 1487ndash1498 httpsdoiorg101111geb12939

Casady GM and SE Marsh 2010 Broad-scale environmental conditionsresponsible for post-fire vegetation dynamics Remote Sensing 2 (12) 2643ndash2664 httpsdoiorg103390rs2122643

Casillo J C Kunst and M Semmartin 2012 Effects of fire and water availabilityon the emergence and recruitment of grasses forbs and woody species in asemiarid Chaco savanna Austral Ecology 37 452ndash459 httpsdoiorg101111j1442-9993201102306x

Cavallero L DR Loacutepez E Raffaele and MA Aizen 2015 Structural-functionalapproach to identify post-disturbance recovery indicators in forests fromnorthwestern Patagonia a tool to prevent state transitions EcologicalIndicators 52 85ndash95 httpsdoiorg101016jecolind201411019

Cavallero L E Raffaele and MA Aizen 2013 Birds as mediators of passiverestoration during early post-fire recovery Biological Conservation 158 342ndash350 httpsdoiorg101016jbiocon201210004

Certini G 2005 Effects of fire on properties of forest soils a review Oecologia143 (1) 1ndash10 httpsdoiorg101007s00442-004-1788-8

Chaneton EJ CN Maziacutea M Machera A Uchitel and CM Ghersa 2004Establishment of honey locust (Gleditsia triacanthos) in burned Pampeangrasslands Weed Technology 18 1325ndash1329 httpsdoiorg1016140890-037X(2004)018[1325EOHLGT]20CO2

Chapin III FS PA Matson and PM Vitousek 2011 Principles of terrestrialecosystem ecology New York Springer Science amp Business Media httpsdoiorg101007978-1-4419-9504-9

Cochrane MA 2003 Fire science for rainforests Nature 421 913ndash919 httpsdoiorg101038nature01437

Cohn JS J Di Stefano F Christie G Cheers and A York 2015 How doheterogeneity in vegetation types and post-fire age-classes contribute toplant diversity at the landscape scale Forest Ecology and Management 34622ndash30 httpsdoiorg101016jforeco201502023

Giorgis et al Fire Ecology (2021) 1711 Page 16 of 20

Coop JD RT Massatti and AW Schoettle 2010 Subalpine vegetation patternthree decades after stand-replacing fire effects of landscape context andtopography on plant community composition tree regeneration anddiversity Journal of Vegetation Science 21 (3) 472ndash487 httpsdoiorg101111j1654-1103200901154x

Coop JD SA Parks CS Stevens-Rumann SD Crausbay PE Higuera MDHurteau A Tepley E Whitman T Assal BM Collins KT Davis S DobrowskiDA Falk PJ Fornwalt PZ Fuleacute BJ Harvey VR Kane CE Littlefield EQMargolis M North M-A Parisien S Prichard and KC Rodman 2020Wildfire-driven forest conversion in Western North American landscapesBioscience 70 (8) 659ndash673 httpsdoiorg101093bioscibiaa061

Defosseacute GE MM Godoy LO Bianchi NS Lederer and C Kunst 2015 Fire historyfire ecology and management in Argentine Patagonia from ancient times tonowadays In Current international perspectives on wildland fires mankind and theenvironment ed P Leblon 177ndash210 New York Nova Science Publishers

Di Bella CM EG Jobbaacutegy JM Paruelo and S Pinnock 2006 Continental firedensity patterns in South America Global Ecology amp Biogeography 15 (2)192ndash199 httpsdoiorg101111j1466-822X200600225x

Dirzo R HS Young M Galetti G Ceballos NJB Isaac and B Collen 2014Defaunation in the Anthropocene Science 345 (6195) 401ndash406 httpsdoiorg101126science1251817

Doherty TS EJB van Etten RA Davis C Knuckey JQ Radford and SADalgleish 2017 Ecosystem Responses to Fire Identifying Cross-taxa Contrastsand Complementarities to Inform Management Strategies Ecosystems 20872ndash884 httpsdoiorg101007s10021-016-0082-z

Durigan G NAL Pilon RCR Abreu WA Hoffmann M Martins BF Fiorillo AZAntunes AP Carmignotto JB Maravalhas J Vieira and HL Vasconcelos2020 No net loss of species diversity after prescribed fires in the Braziliansavanna Frontiers in Forests and Global Change 19 February 2020 httpsdoiorg103389ffgc202000013

Dwyer E S Pinnock JM Greacutegoire and JMC Pereira 2000 Global spatial andtemporal distribution of vegetation fire as determined from satelliteobservations International Journal of Remote Sensing 21 1289ndash1302 httpsdoiorg101080014311600210182

Farnsworth LM DG Nimmo LT Kelly AF Bennett and MF Clarke 2014 Doespyrodiversity beget alpha beta or gamma diversity A case study usingreptiles from semi-arid Australia Diversity and Distributions 20 (6) 663ndash673httpsdoiorg101111ddi12181

Fontaine JB and PL Kennedy 2012 Meta-analysis of avian and small-mammalresponse to fire severity and fire surrogate treatments in US fire-prone forestsEcological Applications 2 (5) 1547ndash1561 httpsdoiorg10189012-00091

Galiacutendez G F Biganzoli P Ortega-Baes and AL Scopel 2009 Fire responses ofthree co-occurring Asteraceae shrubs in a temperate savanna in South AmericaPlant Ecology 202 (1) 149ndash158 httpsdoiorg101007s11258-008-9537-4

Garciacutea Y MC Castellanos and JG Pausas 2016 Fires can benefit plants bydisrupting antagonistic interactions Oecologia 182 1165ndash1173 httpsdoiorg101007s00442-016-3733-z

Geary WL DG Nimmo TS Doherty EG Ritchie and AIT Tulloch 2019 Threatwebs Reframing the co-occurrence and interactions of threats tobiodiversity Journal of Applied Ecology 56 1992ndash1997 httpsdoiorg1011111365-266413427

Gerwing JJ 2002 Degradation of forests through logging and fire in the easternBrazilian Amazon Forest Ecology and Management 157 131ndash141 httpsdoiorg101016S0378-1127(00)00644-7

Giorgis M AM Cingolani and M Cabido 2013 El efecto del fuego y lascaracteriacutesticas topograacuteficas sobre la vegetacioacuten y las propiedades del sueloen la zona de transicioacuten entre bosques y pastizales de las sierras de CoacuterdobaArgentina Boletiacuten la Sociedad Argentina Botaacutenica 48 (3-4) 493ndash513 [inSpanish] httpsdoiorg103105518512372v48n3-47555

Griffiths AD and BW Brook 2014 Effect of fire on small mammals a systematicreview International Journal of Wildland Fire 23 (7) 1034ndash1043 httpsdoiorg101071WF14026

Gurevitch J GA Fox GM Wardle and D Taub 2011 Emergent insights fromthe synthesis of conceptual frameworks for biological invasions EcologyLetters 14 407ndash418

Gurevitch J PS Curtis and MH Jones 2001 Meta-analysis in ecologyAdvances in Ecological Research 32 199ndash247 httpsdoiorg101016S0065-2504(01)32013-5

Gurevitch J and LV Hedges 1999 Statistical issues in ecological meta-analysesEcology 80 (4) 1142ndash1149 httpsdoiorg1018900012-9658(1999)080[1142SIIEMA]20CO2

Gurvich DE L Enrico and AM Cingolani 2005 Linking plant functional traitswith post fire sprouting vigour in woody species in central Argentina AustralEcology 30 789ndash796 httpsdoiorg101111j1442-9993200501522x

Harris RMB TA Remenyi GJ Williamson NL Bindoff and DMGS Bowman2016 Climate-vegetation-fire interactions and feedbacks trivial detail ormajor barrier to projecting the future of the Earth system WIRES ClimateChange 7 (6) 910ndash931 httpsdoiorg101002wcc428

He T BB Lamont and JG Pausas 2019 Fire as a key driver of Earthrsquos biodiversityBiological Reviews 94 (6) 1983ndash2010 httpsdoiorg101111brv12544

Hedges LV and I Olkin 2014 Statistical methods for meta-analysis Orlando AcademicHerrero M R Torres and D Renison 2016 Do wildfires promote woody species

invasion in a fire-adapted ecosystem Post-fire resprouting of native andnon-native woody plants in central Argentina Environmental Management 57(2) 308ndash317 httpsdoiorg101007s00267-015-0616-8

Heydari M M Faramarzi and D Pothier 2016 Post-fire recovery of herbaceousspecies composition and diversity and soil quality indicators one year afterwildfire in a semi-arid oak woodland Ecological Engineering 94 688ndash697httpsdoiorg101016jecoleng201605032

Hijmans RJ SE Cameron JL Parra PG Jones and A Jarvis 2005 Very highresolution interpolated climate surfaces for global land areas InternationalJournal of Climatology 25 (15) 1965ndash1978 httpsdoiorg101002joc1276

Hoffmann BD and AN Andersen 2003 Responses of ants to disturbance inAustralia with particular reference to functional groups Austral Ecology 28(4) 444ndash464 httpsdoiorg101046j1442-9993200301301x

Hoffmann WA VMPC Lucatelli FJ Silva INC Azeuedo MdS Marinho AMSAlbuquerque AdO Lopes and SP Moreira 2004 Impact of the invasivealien grass Melinis minutiflora at the savanna-forest ecotone in the BrazilianCerrado Diversity and Distribution 10 (2) 99ndash103 doi httpsdoiorg101111j1366-9516200400063x

Hoffmann WA and AG Moreira 2002 The role of fire in population dynamicsof woody plants In Cerrados Brazil ecology and natural history of aNeotropical savanna ed PS Oliviera and RJ Marquis 159ndash177 New YorkColumbia University Press

Hoffmann WA and OT Solbrig 2003 The role of topkill in the differentialresponse of savanna woody species to fire Forest Ecology and Management180 273ndash286 httpsdoiorg101016S0378-1127(02)00566-2

Jacobsen AL MF Tobin HS Toschi MI Percolla and RB Pratt 2016 Structuraldeterminants of increased susceptibility to dehydration-induced cavitation inpost-fire resprouting chaparral shrubs Plant Cell and Environment 39 (11)2473ndash2485 httpsdoiorg101111pce12802

Jaureguiberry P A Cuchietti LD Gorneacute GA Bertone and S Diacuteaz 2020 Post-fireresprouting capacity of seasonally dry forest species - two quantitative indices ForestEcology and Management 473 118267 httpsdoiorg101016jforeco2020118267

Jennions MD CJ Lortie MS Rosenberg and HR Rothstein 2013Publication and related biases In Handbook of meta-analysis in ecologyand evolution ed J Koricheva J Gurevitch and K Mendersen 207ndash236Princeton Princeton University Press httpsdoiorg1023943princeton97806911372850030014

Keeley JE 2009 Fire intensity fire severity and burn severity a brief review andsuggested usage International Journal of Wildland Fire 18 116ndash126 httpsdoiorg101071WF07049

Keeley JE WJ Bond RA Bradstock JG Pausas and PW Rundel 2012 Fire inMediterranean ecosystems ecology evolution and management CambridgeCambridge University Press httpsdoiorg101017CBO9781139033091

Keeley JE CJ Fotheringham and M Baer-Keeley 2005 Determinants of postfirerecovery and succession in Mediterranean-climate shrublands of CaliforniaEcological Applications 15 (5) 1515ndash1534 httpsdoiorg10189004-1005

Keeley JE JG Pausas PW Rundel WJ Bond and RA Bradstock 2011 Fire asan evolutionary pressure shaping plant traits Trends in Plant Science 16 406ndash411 httpsdoiorg101016jtplants201104002

Keeley SC JE Keeley SM Hutchinson and AW Johnson 1981 Postfiresuccession of the herbaceous flora in Southern California chaparral Ecology62 (6) 1608ndash1621 httpsdoiorg1023071941516

Kelly LT L Brotons KM Giljohann MA McCarthy JG Pausas and AL Smith 2018Bridging the divide integrating animal and plant paradigms to secure the future ofbiodiversity in fire-prone ecosystems Fire 1 29 httpsdoiorg103390fire1020029

Knicker H 2007 How does fire affect the nature and stability of soil organicnitrogen and carbon A review Biogeochemistry 85 (1) 91ndash118 httpsdoiorg101007s10533-007-9104-4

Knoechelmann CM and HC Morais 2008 Visitas de formigas (HymenopteraFormicidae) a nectaacuterios extra-florais de Stryphnodendron adstringens (Mart)

Giorgis et al Fire Ecology (2021) 1711 Page 17 of 20

Cov (Fabaceae Mimosoideae) em uma aacuterea de cerrado frequumlentementequeimado Revista Brasileira de Zoociecircncias 10 1 [in Portuguese]

Kowaljow E MS Morales JL Whitworth-Hulse SR Zeballos MA Giorgis MRCatoacuten and DE Gurvich 2019 A 55-year-old natural experiment givesevidence of the effects of changes in fire frequency on ecosystem propertiesin a seasonal subtropical dry forest Land Degradation and Development 30(3) 266ndash277 httpsdoiorg101002ldr3219

Kral KC RF Limb JP Harmon and TJ Hovick 2017 Arthropods and fireprevious research shaping future conservation Rangeland Ecology andManagement 70 (5) 589ndash598 httpsdoiorg101016jrama201703006

Krawchuk MA and MA Moritz 2011 Constraints on global fire activity vary acrossa resource gradient Ecology 92 121ndash132 httpsdoiorg10189009-18431

Kunst C S Bravo J Panigatti et al eds 2003 Fuego en los ecosistemasArgentinos Santiago del Estero Ediciones INTA [in Spanish]

Kunst C R Ledesma S Bravo et al 2015 Fire history fire ecology andmanagement in the Argentine Chaco In Current international perspectives onwildland fires mankind and the environment ed B Leblon and M Alexander Hauppauge Nova Sciences PublishersChapter 8

Kurten EL 2013 Cascading effects of contemporaneous defaunation on tropicalforest communities Biological Conservation 163 22ndash32 httpsdoiorg101016jbiocon201304025

Lal R 2004 Soil carbon sequestration impacts on global climate change andfood security Science 304 (5677) 5677 pp 1623-1627 httpsdoiorg101126science1097396

Lazarina M J Devalez L Neokosmidis et al 2019 Moderate fire severity is bestfor the diversity of most of the pollinator guilds in Mediterranean pineforests Ecology 100 e02615 httpsdoiorg101002ecy2615

Lee DE 2018 Spotted Owls and forest fire a systematic review and meta-analysis of the evidence Ecosphere 9 (7) e02354 httpsdoiorg101002ecs22354

Lehmann CER TM Anderson M Sankaran et al 2014 Savanna vegetation-fire-climate relationships differ among continents Science 343 (6170) 548ndash552httpsdoiorg101126science1247355

Longo MS C Urcelay and E Nouhra 2011 Long term effects of fire onectomycorrhizas and soil properties in Nothofagus pumilio forests inArgentina Forest Ecology and Management 262 (3) 348ndash354 httpsdoiorg101016jforeco201103041

Longo S E Nouhra BT Goto RL Berbara and C Urcelay 2014 Effects of fire onarbuscular mycorrhizal fungi in the Mountain Chaco Forest Forest Ecologyand Management 315 86ndash94 httpsdoiorg101016jforeco201312027

Lorenz TJ KT Vierling TR Johnson and PC Fischer 2015 The role of woodhardness in limiting nest site selection in avian cavity excavators EcologicalApplications 25 (4) 1016ndash1033 httpsdoiorg10189014-10421

Maestre FT DJ Eldridge and S Soliveres 2016 A multifaceted view on theimpacts of shrub encroachment Applied Vegetation Science 19 (3) 369ndash370httpsdoiorg101111avsc12254

Mandle L J Bufford I Schmidt and C Daehler 2011 Woody exotic plantinvasions and fire reciprocal impacts and consequences for nativeecosystems Biological Invasions 13 1815ndash1827 httpsdoiorg101007s10530-011-0001-3

Marcora PI AE Ferreras SR Zeballos G Funes S Longo C Urcelay and PATecco 2018 Context-dependent effects of fire and browsing on woody alieninvasion in mountain ecosystems Oecologia 188 (2) 479ndash490 httpsdoiorg101007s00442-018-4227-y

Mataix-Solera J A Cerdagrave V Arcenegui A Jordaacuten and LM Zavala 2011 Fireeffects on soil aggregation a review Earth-Science Reviews 109P 44ndash60httpsdoiorg101016jearscirev201108002

Maziacutea CN EJ Chaneton M Machera A Uchitel MV Feler and CM Ghersa2010 Antagonistic effects of large- and small-scale disturbances on exotictree invasion in a native tussock grassland relict Biological Invasions 12 (9)3109ndash3122 httpsdoiorg101007s10530-010-9702-2

McLauchlan KK PE Higuera J Miesel BM Rogers J Schweitzer JK Shuman AJ Tepley JM Varner TT Veblen SA Adalsteinsson JK Balch P Baker EBatllori E Bigio P Brando M Cattau ML Chipman J Coen R Crandall LDaniels N Enright WS Gross BJ Harvey JA Hatten S Hermann REHewitt LN Kobziar JB Landesmann MM Loranty SY Maezumi L MearnsM Moritz JA Myers JG Pausas AFA Pellegrini WJ Platt J Roozeboom HSafford F Santos RM Scheller RL Sherriff KG Smith MD Smith and ACWatts 2020 Fire as a fundamental ecological process research advances andfrontiers Journal of Ecology 108 (5) 2047ndash2069 httpsdoiorg1011111365-274513403

Mestre LAM MA Cochrane and J Barlow 2013 Long-term changes in birdcommunities after wildfires in the central Brazilian Amazon Biotropica 45 (4)480ndash488 httpsdoiorg101111btp12026

Miller RF JC Chambers DA Pyke FB Pierson and C Jason Williams 2013 Areview of fire effects on vegetation and soils in the great basin regionresponse and ecological site characteristics In USDA Forest Service GeneralTechnical Report RMRS-GTR-308 Fort Collins USDA Forest Service RockyMountain Research Station httpsdoiorg102737RMRS-GTR-308

Miller RG R Tangney NJ Enright JB Fontaine DJ Merritt MKJ Ooi KXRuthrof and BP Miller 2019 Mechanisms of fire seasonality effects on plantpopulations Trends in Ecology amp Evolution 34 (12) 1104ndash1117 httpsdoiorg101016jtree201907009

Miranda HS MMC Bustamante and AC Miranda 2002 The fire factor In Thecerrados of Brazil ecology and natural history of a neotropical savanna ed POlibeira and R Marquis 51ndash68 New York Columbia University Press httpsdoiorg107312oliv12042-005

Morales AM N Politi LO Rivera CG Vivanco and GE Defosseacute 2020 Fire anddistance from unburned forest influence bird assemblages in SouthernAndean Yungas of Northwest Argentina a case study Fire Ecology 16 15httpsdoiorg101186s42408-020-00074-0

Moretti M F De Bello SPM Roberts and SG Potts 2009 Taxonomical vsfunctional responses of bee communities to fire in two contrastingclimatic regions Journal of Animal Ecology 78 98ndash108 httpsdoiorg101111j1365-2656200801462x

Murphy BP and J Russell-Smith 2010 Fire severity in a northern Australiansavanna landscape the importance of time since previous fire InternationalJournal of Wildland Fire 19 46ndash51 httpsdoiorg101071WF08202

Murphy EC and WA Lehnhausen 1998 Density and foraging ecology ofwoodpeckers following a stand-replacement fire Journal of WildlifeManagement 62 (4) 1359ndash1372 httpsdoiorg1023073802002

Nakagawa S and ESE Santos 2012 Methodological issues and advances inbiological meta-analysis Evolutionary Ecology 26 1253ndash1274 httpsdoiorg101007s10682-012-9555-5

Nappi A P Drapeau J Giroux and JL Savard 2003 Snag use by foraging black-backed woodpeckers (Picoides arcticus) in a recently burned eastern borealforest Auk 120 (2) 505ndash511 httpsdoiorg101093auk1202505

Nelson ZJ PJ Weisberg and SG Kitchen 2014 Influence of climate andenvironment on post-fire recovery of mountain big sagebrush InternationalJournal of Wildland Fire 23 131ndash142 httpsdoiorg101071WF13012

Nogueira JMP S Rambal JPRAD Barbosa and F Mouillot 2017 Spatialpattern of the seasonal droughtburned area relationship acrossBrazilian biomes sensitivity to drought metrics and global remote-sensing fire products Climate 5 (2) 42 httpsdoiorg103390cli5020042

OrsquoConnor RC JH Taylor and JB Nippert 2020 Browsing and fire decreasesdominance of a resprouting shrub in woody encroached grassland Ecology101 (2) e02935 httpsdoiorg101002ecy2935

Olson DM E Dinerstein ED Wikramanayake ND Burgess GVN Powell ECUnderwood JA DrsquoAmico I Itoua HE Strand JC Morrison CJ Loucks TFAllnutt TH Ricketts Y Kura JF Lamoreux WW Wettengel P Hedao and KR Kassem 2001 Terrestrial ecoregions of the world a new map of life onEarth a new global map of terrestrial ecoregions provides an innovative toolfor conserving biodiversity BioScience 51 933ndash938 httpsdoiorg1016410006-3568(2001)051[0933TEOTWA]20CO2

Parisien M-A and MA Moritz 2009 Environmental controls on the distributionof wildfire at multiple spatial scales Ecological Monographs 79 (1) 127ndash154httpsdoiorg10189007-12891

Pausas J and JE Keeley 2009 A burning story the role of fire in the history oflife Bioscience 59 (7) 593ndash601 httpsdoiorg101525bio200959710

Pausas JG 2019 Generalized fire response strategies in plants and animals Oikos128 147ndash153 httpsdoiorg101111oik05907

Pausas JG and RA Bradstock 2007 Fire persistence traits of plants along aproductivity and disturbance gradient in mediterranean shrublands of south-east Australia Global Ecology and Biogeography 16 (3) 330ndash340 httpsdoiorg101111j1466-8238200600283x

Pausas JG and V de L Dantas 2017 Scale matters fire-vegetation feedbacksare needed to explain tropical tree cover at the local scale Global Ecologyand Biogeography 26 (4) 395ndash399 httpsdoiorg101111geb12562

Pausas JG and JE Keeley 2014 Evolutionary ecology of resprouting andseeding in fire-prone ecosystems New Phytology 204 55ndash65 httpsdoiorg101111nph12921

Giorgis et al Fire Ecology (2021) 1711 Page 18 of 20

Pausas JG and S Paula 2012 Fuel shapes the fire-climate relationship evidencefrom Mediterranean ecosystems Global Ecology and Biogeography 21 (11)1074ndash1082 httpsdoiorg101111j1466-8238201200769x

Pausas JG and E Ribeiro 2013 The global fire-productivity relationship GlobalEcology and Biogeography 22 (6) 728ndash736 httpsdoiorg101111geb12043

Pausas JG and E Ribeiro 2017 Fire and plant diversity at the global scaleGlobal Ecology and Biogeography 26 (8) 889ndash897 httpsdoiorg101111geb12596

Pellegrini AFA A Ahlstroumlm SE Hobbie PB Reich LP Nieradzik AC Staver BCScharenbroch A Jumpponen WRL Anderegg JT Randerson and RBJackson 2018 Fire frequency drives decadal changes in soil carbon andnitrogen and ecosystem productivity Nature 553 (7687) 194ndash198 httpsdoiorg101038nature24668

Pellegrini AFA SE Hobbie and PB Reich 2020a Repeated fire shifts carbonand nitrogen cycling by changing plant inputs and soil decompositionacross ecosystems Ecological Monographs 90 (4) e01409 httpsdoiorg101002ecm1409

Pellegrini AFA KK McLauchlan SE Hobbie MC Mack AL Marcotte DMNelson SS Perakis PB Reich and W Kyle Whittinghill 2020b Frequentburning causes large losses of carbon from deep soil layers in a temperatesavanna Journal of Ecology 8 (4) 1426ndash1441 httpsdoiorg1011111365-274513351

Peacuterez-Meacutendez N P Jordano C Garciacutea and A Valido 2016 The signatures ofAnthropocene defaunation cascading effects of the seed dispersal collapseScientific Reports 6 24820 httpsdoiorg101038srep24820

Pilon NAL MGB Cava WA Hoffmann RCR Abreu A Fidelis and GDurigan 2021 The diversity of post-fire regeneration strategies in thecerrado ground layer Journal of Ecology 109 154ndash166 httpsdoiorg1011111365-274513456

Pilon NAL WA Hoffmann RCR Abreu and G Durigan 2018 Quantifying theshort-term flowering after fire in some plant communities of a cerradograssland Plant Ecology and Diversity 11 259ndash266 httpsdoiorg1010801755087420181517396

Pratt RB AL Jacobsen AR Ramirez AM Helms CA Traugh MF Tobin MSHeffner and SD Davis 2014 Mortality of resprouting chaparral shrubs after afire and during a record drought physiological mechanisms anddemographic consequences Global Change Biology 20 (3) 893ndash907 httpsdoiorg101111gcb12477

Pressler Y JC Moore and MF Cotrufo 2019 Belowground communityresponses to fire meta-analysis reveals contrasting responses of soilmicroorganisms and mesofauna Oikos 128 309ndash327 httpsdoiorg101111oik05738

Prichard SJ CS Stevens-Rumann and PF Hessburg 2017 Tamm reviewshifting global fire regimes lessons from reburns and research needs ForestEcology and Management 396 217ndash233 httpsdoiorg101016jforeco201703035

Pyšek P PE Hulme D Simberloff S Bacher TM Blackburn JT Carlton WDawson F Essl LC Foxcroft P Genovesi JM Jeschke I Kuumlhn AM LiebholdNE Mandrak LA Meyerson A Pauchard J Pergl HE Roy H Seebens Mvan Kleunen M Vilagrave MJ Wingfield and DM Richardson 2020 Scientistsrsquowarning on invasive alien species Biological Review 95 (6) 1511ndash1534httpsdoiorg101111brv12627

R Core Team 2019 R a language and environment for statistical computingVienna R Foundation for Statistical Computing

Raffaele E MA Nuntildeez J Enestroumlm and M Blackhall 2016 Fire as mediator ofpine invasion evidence from Patagonia Argentina Biological Invasions 18597ndash601 httpsdoiorg101007s10530-015-1038-5

Robinson NM SWJ Leonard EG Ritchie M Bassett EK Chia S Buckingham HGibb AF Bennett and MF Clarke 2013 Refuges for fauna in fire-pronelandscapes their ecological function and importance Journal of AppliedEcology 50 1321ndash1329 httpsdoiorg1011111365-266412153

Romano N and N Ursino 2020 Forest fire regime in a Mediterraneanecosystem unraveling the mutual interrelations between rainfall seasonalitysoil moisture drought persistence and biomass dynamics Fire 3 49 httpsdoiorg103390fire3030049

Rostagno CM GE Defosseacute and HF Del Valle 2006 Postfire vegetationdynamics in three rangelands of Northeastern Patagonia ArgentinaRangeland Ecology and Management 59 (2) 163ndash170 httpsdoiorg10211105-020R11

Schepps J S Lohr and T Martin 1999 Does tree hardness influence nest tree selectionby excavating bird species Auk 116 658ndash665 httpsdoiorg1023074089327

Scholes RJ and SR Archer 1997 Tree-grass interactions in savannas AnnualReview of Ecology and Systematics 28 517ndash544 httpsdoiorg101146annurevecolsys281517

Silva LCR WA Hoffmann DR Rossatto M Haridasan AC Franco and WRHorwath 2013 Can savannas become forests A coupled analysis of nutrientstocks and fire thresholds in central Brazil Plant and Soil 373 (1-2) 829ndash842httpsdoiorg101007s11104-013-1822-x

Silveira JM J Barlow J Louzada and P Moutinho 2010 Factors affecting theabundance of leaf-litter arthropods in unburned and thrice-burnedseasonally-dry Amazonian forests PLoS One 5 (9) 1ndash7 httpsdoiorg101371journalpone0012877

Silveira JM J Louzada J Barlow R Andrade L Mestre R Solar S Lacau and MA Cochrane 2016 A multi-taxa assessment of biodiversity change aftersingle and recurrent wildfires in a Brazilian Amazon forest Biotropica 48 170ndash180 httpsdoiorg101111btp12267

Souchie FF JRR Pinto E Lenza L Gomes L Maracahipes-Santos and DVSilveacuterio 2017 Post-fire resprouting strategies of woody vegetation in theBrazilian savanna Acta Botanica Brasilica 31 (2) 260ndash266 httpsdoiorg1015900102-33062016abb0376

Staver AC S Archibald and SA Levin 2011 The Global extent anddeterminants of savanna and forest as alternative biome states Science 334(6053) 230ndash232 httpsdoiorg101126science1210465

Torres RC MA Giorgis C Trillo et al 2014 Post-fire recovery occursoverwhelmingly by resprouting in the Chaco Serrano forest of CentralArgentina Austral Ecology 39 346ndash354 httpsdoiorg101111aec12084

Tuck SL C Winqvist F Mota J Ahnstroumlm LA Turnbull and J Bengtsson 2014Land-use intensity and the effects of organic farming on biodiversity ahierarchical meta-analysis Journal of Applied Ecology 51 (3) 746ndash755 httpsdoiorg1011111365-266412219

Tummers B 2006 DataThief III v11 Available from httpwwwdatathieforgUehara-Prado M A de M Bello J de O Fernandes AJ Santos IA Silva and M

V Cianciaruso 2010 Abundance of epigaeic arthropods in a Braziliansavanna under different fire frequencies Zoologia 27 (5) 718ndash724 httpsdoiorg101590S1984-46702010000500008

Veblen TT T Kitzberger E Raffaele and DC Lorenz 2003 Fire history andvegetation changes in northern Patagonia Argentina BT In Fire and climaticchange in temperate ecosystems of the western Americas ed TT Veblen WLBaker G Montenegro and TW Swetnam 265ndash295 New York Springerhttpsdoiorg1010070-387-21710-X_9

Veblen TT T Kitzberger E Raffaele M Mermoz ME Gonzaacutelez JS Sibold and AHolz 2008 The historical range of variability of fires in the AndeanPatagonian Nothofagus forest region International Journal of Wildland Fire 17(6) 724ndash741 httpsdoiorg101071WF07152

Vidaller C T Dutoit H Ramone and A Bischoff 2019 Fire increases thereproduction of the dominant grass Brachypodium retusum andMediterranean steppe diversity in a combined burning and grazingexperiment Applied Vegetation Science 22 127ndash137 httpsdoiorg101111avsc12418

Viechtbauer W 2010 Conducting meta-analyses in R with the metaforpackage Journal of Statistical Software 36 (3) 1ndash48 httpsdoiorg1018637jssv036i03

Violle C M Navas D Vile and E Kazakou 2007 Let the concept of trait befunctional Oikos 116 (5) 882ndash892 httpsdoiorg101111j0030-1299200715559x

Vitousek PM and RW Howarth 1991 Nitrogen limitation on land and in thesea how can it occur Biogeochemistry 13 (2) 87ndash115 httpsdoiorg101007BF00002772

Whelan RJ L Rodgerson CR Dickman and EF Sutherland 2002 Critical lifecycles of plants and animals developing a process-based understanding ofpopulation changes in fire-prone landscapes In Flammable Australia the fireregimes and biodiversty of a continent ed RA Bradstock JE Williams and AM Gill 94ndash124 Cambridge Cambridge University Press

Whittaker RH 1975 Communities and ecosystems 2nd ed New York MacmillanPublishing

Whittaker RH and PI Marks 1975 Methods of assessing terrestrialproductivity In Primary productivity of the biosphere ed H Lieth and RHWhittaker 55ndash118 New York Springer httpsdoiorg101007978-3-642-80913-2_4

Winkler H and DA Christie 2002 Family Picidae (woodpeckers) In Handbook ofthe birds of the world volume 7 Jacamars to Woodpeckers ed J del Hoyo AElliott and J Sargatal 296ndash555 Barcelona Lynx Edicions

Giorgis et al Fire Ecology (2021) 1711 Page 19 of 20

Zeballos SR MA Giorgis MR Cabido ATR Acosta M del Rosario Iglesiasand JJ Cantero 2020 The lowland seasonally dry subtropical forests incentral Argentina vegetation types and a call for conservationVegetation Classification and Survey 1 (1) 87ndash102 httpsdoiorg103897VCS202038013

Publisherrsquos NoteSpringer Nature remains neutral with regard to jurisdictional claims inpublished maps and institutional affiliations

Giorgis et al Fire Ecology (2021) 1711 Page 20 of 20