Cipo Canastero (Asthenes luizae): revise global status?

BirdLife species factsheet for Cipo Canastero

This discussion was first published as part of the 2019 Red List update. At the time a decision regarding the status of this species was pended and the post remained open. Following experts’ review of the species factsheet, the topic has now been updated to reflect the most recent information. The initial topic can be found here.

Cipo Canastero (Asthenes luizae) occurs in the central-southern Espinhaço Range, Minas Gerais, Brazil. It was discovered in 1985 in a small area within the Serra do Cipó (Pearman 1990, Vielliard 1990). Subsequent surveys detected the species in several other ridges within the central Espinhaço Range (Chaves Cordeiro et al. 1998, Vasconcelos et al. 2002, Vasconcelos and D’Angelo Neto 2007, Vasconcelos 2009, Chaves 2014).

Cipo Canastero is restricted to campo rupestre: isolated rocky outcrops with dry, bushy vegetation. It ranges from 1,100 to 2,000 m (Freitas et al. 2019) and prefers steep, well vegetated slopes. Currently, the most important threat to the species is thought to be brood parasitism by the recently introduced Shiny Cowbird (Molothrus bonariensis) (Costa 2015). Other threats include fires caused by tourist campfires or from escaped agricultural fires (L. E. Lopes in litt. 2009, Vasconceles 2009, Costa 2015), grazing, incursion of invasive grasses (Freitas et al. 2019), development for tourism and infrastructure, mineral extraction and climate change. A model of the impact of climate on the potential range of rupestrian grasslands in the Espinhaço Range predicted that around half of these grasslands would be lost by 2080 (Bitencourt et al. 2016). Other models indicated that rupestrian grasslands might lose up to 95% of their current suitable area by the end of the century (Wilson Fernandes et al. 2014).

Currently, Cipo Canastero is listed as Near Threatened, approaching the threshold for listing as threatened under Criterion B1ab(iii) (BirdLife International 2020). However, this is no longer tenable because this was based on an Extent of Occurrence (EOO) value calculated as the ‘area of mapped range’. This is no longer appropriate, and the EOO should be calculated using a Minimum Convex Polygon (see IUCN 2001, 2012, Joppa et al. 2016), as EOO is a measure of the spatial spread of areas occupied by a species, not the actual area it occupies. After re-calculating the EOO for Cipo Canastero, the species appears to warrant a change in Red List status. Therefore, we present here our reassessment against all criteria for the species:

Criterion A – Studies of ringed individuals and territories in Serra do Cipó over six years by Costa (2015) indicated a declining population trend from 2009 to 2015. Therefore, there is an inferred continuing decline in the number of mature individuals. The species’s generation length is estimated to be 2.89 years (Bird et al. 2020)*; since three generations would be less than ten years for this species, reductions under Criterion A should be assessed over a period of ten years. There is no quantified estimate of the rate of reduction over the past ten years, or ten years into the future, so the species cannot be assessed under Criterion A.

Criterion B – The newly calculated Extent of Occurrence (EOO) for this species, based on an updated range map (not yet available on the BirdLife Data Zone), is 37,000 km2. This does not meet or approach the threshold for Vulnerable (EOO < 20,000 km2). Therefore, Cipo Canstero is assessed as Least Concern under Criterion B1.

Based on the total area of 4 km2 squares intersecting the species’s occurrence records, the minimum Area of Occupancy (AOO) was estimated to be 228 km2 (Freitas et al. 2019). However, the species is unlikely to be confined only to the 4 km2 squares in which it has been recorded, so this likely represents an underestimate.

Based on the total area of 4 km2 squares intersecting the species’s mapped range, the maximum AOO is inferred to be 17,100 km2. However, the species has specialist habitat requirements and is unlikely to occur everywhere in the mapped range, so this figure likely represents an overestimate.

The best estimate of the AOO is inferred to be 6,000 km2,  based on the total area of 4 km2 squares intersecting the area of a species distribution model (Freitas et al. 2019), with areas without records removed.

This range of estimates of AOO could qualify the species for Endangered or Least Concern under Criterion B2. The best estimate does not meet or approach the threshold for listing the species as threatened under Criterion B2 and would qualify the species for listing as Least Concern under Criterion B2. If there is considerable uncertainty around whether the true AOO falls beneath the threshold for listing the species as threatened under this criterion, the species could qualify for Near Threatened. To list the species as threatened or Near Threatened under Criterion B, at least two further conditions have to be met.

The species is not severely fragmented, since >50% of its total area of occupancy is not in habitat patches that are either smaller than would be required to support a viable population, or separated from other habitat patches by a large distance. According to the IUCN Red List Categories and Criteria, ‘the term ‘location’ defines a geographically or ecologically distinct area in which a single threatening event can rapidly affect all individuals of the taxon present’ (IUCN 2012). The main threat to the species is considered to be brood parasitism by the recently introduced Shiny Cowbird (Molothrus bonariensis) (Costa 2015). Although this threat is thought to be causing population declines across the species’s range, the number of locations would likely be greater than ten. Another threat is the risk of fires caused by tourist campfires, or from escaped agricultural fires (L. E. Lopes in litt. 2009, Vasconceles 2009, Costa 2015). The species occurs on nine ‘sky islands’, but it is likely that it would take more than one fire to destroy all habitat at some of these (particularly the Diamantina Plateau), so the number of locations based on the threat of fire would be greater than ten. A model of the impact of climate change on the potential range of rupestrian grassland in the Espinhaço Range predicted that around half would be lost by 2080 (Bitencourt et al. 2016). Other models indicated that rupestrian grasslands might lose up to 95% of their current suitable area by the end of the century (Wilson Fernandes et al. 2014). Although climate change is projected to impact the majority of the species’s range, this is uncertain and is predicted to happen over a period of decades, so is unlikely to rapidly (i.e. within one generation) affect all individuals over a large area. Therefore, we consider that the number of locations is likely to be greater than ten, but based on the threat of brood parasitism by Shiny Cowbird, it may be approaching ten.

Studies of ringed individuals and territories in Serra do Cipó over six years by Costa (2015) indicated a declining population trend from 2009 to 2015. Therefore, there is an inferred continuing decline in the number of mature individuals. A continuing decline in the area and quality of habitat, EOO, AOO and number of locations may also be projected from models of the impact of climate change on rupestre grassland (Wilson Fernandes et al. 2014, Bitencourt et al. 2016). Condition b is met.

There is no evidence that the species’s population or range size are undergoing extreme fluctuations. Condition c is not met.

Although the species’s minimum estimate of AOO would qualify the species for listing as Endangered, the best estimate of AOO does not meet or approach the threshold for Vulnerable. Although condition b is met, condition c is not and we do not have evidence to suggest that condition a is met. Depending on whether the number of locations is thought to be close to ten, Cipo Canastero may be assessed as Near Threatened, approaching the threshold for listing as threatened under Criterion B2ab(i,ii,iii,iv,v), or as Least Concern.

Criterion C – Surveys in 2009-2010 in Serra do Cipó found a mean population density of 29 individuals per km2 in campo rupestre habitat, and 13 individuals and five breeding pairs per km2 across the study area (Freitas 2011). Another analysis of surveys at Serra do Cipó provided mean population densities of 35.2 adult individuals per km2 in campo rupestre habitat, and 13.6 adults per km2 across the study area (Freitas et al. 2019). Considering paired individuals only, this study found an average of 22.9 per km2 in campo rupestre habitat, and 8.7 per km2 across the study area (Freitas et al. 2019). Surveys from 2009-2015 in Serra do Cipó found an average density of 17 adult individuals per km2 of rocky outcrops, and 8 adults per km2 of study area, with a lower density of paired individuals (7 and 3 respectively) due to the sex ratio bias (Costa 2015).

Based on the area of habitat within the Cipo Canastero’s range, the global population size was projected to be around 21,800 (7,400 – 47,600) paired individuals (Costa 2015). The species has a strongly male-biased sex ratio (Freitas 2011, Costa 2015), with an adult population composed of c.70% males (Costa 2015, Freitas et al. 2019). Accordingly, the population size estimate was based on the number of paired individuals (Costa 2015). This range of estimates could qualify the species as Vulnerable, Near Threatened or Least Concern under Criterion C. The best estimate, 21,800 mature individuals, would qualify the species for Least Concern under Criterion C.

To list the species as threatened or Near Threatened under Criterion C, further conditions must also be met. Studies of ringed individuals and territories in Serra do Cipó over six years by Costa (2015) indicated a declining population trend from 2009 to 2015. Therefore, there is an inferred continuing decline in the number of mature individuals. We do not have data from which to estimate the rate of decline, so the species cannot be assessed under Criterion C1.

The distribution range is divided into three areas separated by some distance and there are at least three subpopulations. Costa (2015) estimated 700 mature individuals in the northern part of the range, 1,900 in the central part and 19,800 in the southern part. However, it is likely that there are further subpopulation divisions within these blocks, particularly the southern block, due to narrower gaps of around 10 km between areas of suitable elevation. The species is thought to have a low dispersal distance (Costa 2015). A study of the species’s habitat and elevational range identified nine ‘sky islands’ within the range, which may represent separate subpopulations (Freitas et al. 2019). The largest of these is the Diamantina Plateau, which was estimated to represent 55% of the total range (Freitas et al. 2019). Based on the population estimates described above, and assuming that the Diamantina Plateau supports around 55% of the total population, there are likely to be at least 4,000 mature individuals in the largest subpopulation. Neither condition C2a(i) nor condition C2a(ii) is met. There is no evidence that the species’s population size is undergoing extreme fluctuations so the species doesn’t meet condition C2(b).

Although the minimum inferred population size falls beneath the threshold for listing Cipo Canastero as Vulnerable under Criterion C and there is a continuing decline in population size, the best estimate of the population size does not meet or approach the threshold for listing the species as Vulnerable. Furthermore, there is no data from which to estimate or infer a rate of decline, and none of conditions C2a(i), C2a(ii) or C2b are met. The species therefore qualifies as Least Concern under Criterion C.

Criterion D – The population size is too large to approach the threshold for Vulnerable (1,000 mature individuals). Therefore, Cipo Canastero is assessed as Least Concern under Criterion D.

Criterion E – A Population Viability Analysis (PVA) has been carried out using the software VORTEX, and based on data gained through surveys carried out in 2009-2015 in Serra do Cipó (Costa 2015). Twelve candidate models were constructed by varying three parameters (juvenile mortality, sex-specific mortality rates and the rate of reproductive success), and three models were selected from those that best fitted the population trends observed at Serra do Cipó in 2009-2015, representing optimistic, pessimistic and neutral scenarios. These models were then used to estimate the probability of extinction over 12, 20, 50 and 100 years for the global population. The probability of extinction in 12 years was estimated as 0% for the neutral and optimistic scenarios and 0.8 (± 0.4%) for the pessimistic scenario. The probability of extinction in 20 years was estimated as 0% for the optimistic scenario, 58.6% (± 2.2%) for the neutral scenario, and 95.6% (± 0.9%) for the pessimistic scenario. The probability of extinction in 100 years was estimated as 100% for all three scenarios. If accepted, these results would qualify the species for listing as Endangered under Criterion E.

The Red List Guidelines provide strict requirements and recommendations for the use of PVAs in Red List Assessments. In particular, it is noted that ‘when there is not sufficient data, or when the available information is too uncertain, it is risky to make a criterion E assessment with any method, including PVA’ (IUCN Standards and Petitions Committee 2019). In order to decide whether there is sufficient data, the Red List Guidelines recommend that model parameters are estimated, incorporating the uncertainties in the data, to provide ‘optimistic’ and ‘pessimistic’ estimates. These ranges of estimates may then be used to create a range of models to give a range of extinction risk estimates, and the range of these estimates used to determine whether the results are useful, and hence whether there is enough data (see IUCN Standards and Petitions Committee 2019, p. 74). The Red List Guidelines suggest that ‘all parameters be specified as ranges (or as distributions) that reflect uncertainties in the data (lack of knowledge or measurement errors)’ (see IUCN Standards and Petitions Committee 2019, p. 76).

Although a range of twelve candidate scenarios were created by Costa (2015), they did not incorporate the uncertainty of all parameters used, and they were only tested for the period of the population study (2009-2015), with an initial population size of 30 individuals. The optimistic and pessimistic scenarios used to assess extinction risk of the global population were selected out of those that best fitted trends seen at Serra do Cipó over 2009-2015, and did not incorporate the full range of uncertainty in the parameters. Indeed, in the documentation, it is explained that the full standard error in the percentage of breeding females was not included in the model, as it was too large. No range of estimates was included in the model for the initial global population size. The datasets from which the parameters were derived were small, including a study of breeding success based on 17 nests, and a demographic study based on 37 ringed (banded) individuals and monitoring of 22 territories (Costa 2015). The documentation acknowledges that some of the parameter estimates were uncertain due to the small number of individuals studied.

The Red List Guidelines further state that ‘the documentation should include a list of assumptions of the analysis, and provide explanations and justifications for these assumptions’ (see IUCN Standards and Petitions Committee 2019, p. 76). Whilst the documentation provided by Costa (2015) is extensive and explains most of the assumptions and judgements made in the derivation of the parameters, there is insufficient explanation of the initial global population size used in the model. The population size used was the estimated number of individuals in breeding pairs, as described under Criterion C above, and excluded both immature individuals and a proportion of adult males in the population. Although population size under Criteria C and D should be based on the number of mature individuals, this is not a requirement for PVAs used under Criterion E. It is not clear why the initial population size did not include all individuals, and since the model was not also run with all individuals included, we cannot tell how their inclusion would have affected the estimated extinction probabilities.

Lastly, the Red List Guidelines state that ‘all data files that were used in the analysis’ should be provided (see IUCN Standards and Petitions Committee 2019, p. 76). This is not the case here, although the parameter values used were specified.

We therefore consider that there is insufficient information presented for us to be able to determine whether there was enough data to provide useful results in the PVA, and the information that is presented suggests that there may not have been. The species cannot be assessed under Criterion E.

Based on the above assessment, it is proposed to list Cipo Canastero (Asthenes luizae) as Least Concern, or as Near Threatened, approaching the thresholds listing as threatened under Criterion B2ab(ii,iii,iv,v). The final category will depend on the determined AOO and number of locations. To allow us to achieve a clearer assessment of the status of Cipo Canastero, information is requested on the potential impact of threats of the species, and particularly the area (or proportion) of the range that is likely to be rapidly affected by threats.

Spatial files representing the species’s range (such as the Species Distribution Model) are also welcomed. GIS files should be emailed to redlistteam@birdlife.org.

Please note that this topic is not designed to be a general discussion about the ecology of the species, rather a discussion of its Red List status. Therefore, please make sure your comments are relevant to the discussion outlined in the topic. By submitting a comment, you confirm that you agree to the Comment Policy.

*Bird generation lengths are estimated using the methodology of Bird et al. (2020), as applied to parameter values updated for use in each IUCN Red List for birds reassessment cycle. Values used for the current assessment are available on request. We encourage people to contact us with additional or improved values for the following parameters; adult survival (true survival accounting for dispersal derived from an apparently stable population); mean age at first breeding; and maximum longevity (i.e. the biological maximum, hence values from captive individuals are acceptable).

An information booklet on the Red List Categories and Criteria can be downloaded here and the Red List Criteria Summary Sheet can be downloaded here. Detailed guidance on IUCN Red List terms and definitions and the application of the Red List Categories and Criteria can be downloaded here.

References

Bird, J. P.; Martin, R.; Akçakaya, H. R.; Gilroy, J.; Burfield, I. J.; Garnett, S.; Symes, A.; Taylor, J.; Šekercioğlu, Ç.; Butchart, S. H. M. (2020). Generation lengths of the world’s birds and their implications for extinction risk. Conservation Biology online first view.

BirdLife International. 2020. Species factsheet: Asthenes luizae. http://www.birdlife.org.

Bitencourt, C., Rapini, A., Damascena, L.S. and Junior, P.D.M. 2016. The worrying future of the endemic flora of a tropical mountain range under climate change. Flora-Morphology, Distribution, Functional Ecology of Plants 218: 1-10.

Chaves Cordeiro, P. H.; de Melo, T. A., Jr.; Vasconcelos, M. F. de. 1998. A range extension for Cipó Canastero Asthenes luizae in Brazil. Cotinga 10: 64-65.

Chaves, A. V. 2014. Biogegrafia historica e sistematica das aves endemicas dos topos de montanha do leste do Brasil. Departamento De Biologia Geral, Universidade Federal de Minas Gerais.

Costa, L. M. 2015. História natural, demografia, viabilidade populacional e conservação de Asthenes luizae (Furnariidae), ave endêmica dos campos rupestres da Cadeia do Espinhaço, Minas Gerais. Instituto de Ciências Biológicas, Universidade Federal de Minas Gerais.

Fernandes, G. W., Barbosa, N. P., Negreiros, D. and Paglia, A. P. 2014. Challenges for the conservation of vanishing megadiverse rupestrian grasslands. Natureza & Conservação 2(12): 162-165.

Freitas, G.H., Costa, L.M., Chaves, A.V., Vasconcelos, M.F., Ribeiro, L.C., Silva, J.C., Souza, R.A., Santos, F.R. and Rodrigues, M. 2019. Geographic range and conservation of the Cipo Canastero Asthenes luizae, an endemic furnariid of Brazilian sky islands. Bird Conservation International doi:10.1017/S0959270919000418: 1-16.

IUCN Standards and Petitions Committee. 2019. Guidelines for Using the IUCN Red List Categories and Criteria. Version 14. Prepared by the Standards and Petitions Committee. Downloadable from http://www.iucnredlist.org/documents/RedListGuidelines.pdf.

Pearman, M. 1990. Behaviour and vocalizations of an undescribed canastero Asthenes sp. from Brazil. Bulletin of the British Ornithologists’ Club 110: 145-153.

Vasconcelos, M. F. de; D’Angelo Neto, S. 2007. Padrões de distribuição e conservação da avifauna na região central da Cadeia do Espinhaço e áreas adjacentes, Minas Gerais, Brasil. Cotinga 28: 27-44.

Vasconcelos, M. F. de; Neto, S. D.; Rodrigues, M. 2002. A range extension for the Cipó Canastero Asthenes luizae and the consequences for its conservation status. Bulletin of the British Ornithologists’ Club 122: 7-10.

Vielliard, J. 1990. Uma nova espécie de Asthenes da serra do Cipó, Minas Gerais, Brasil. Ararajuba: 121-122.

This entry was posted in Americas, South America and tagged , . Bookmark the permalink.

6 Responses to Cipo Canastero (Asthenes luizae): revise global status?

  1. Lílian M. Costa & Guilherme H. S. Freitas says:

    We emailed GIS files of the species distribution model representing the species’ range used in Freitas et al. (2019a), and we would like to contribute to the discussion adding some comments to what was already exposed here in the forum.

    1) About the EOO and AOO:

    We think that the best estimates of the EOO and AOO for this species are those published in Freitas et al. (2019), because it included the most extensive review of the species records (including previously unpublished data), and it was calculated following the IUCN guidelines for assessment under criterion B; i.e., EOO was calculated as the area within the minimum convex polygon (MCP) constructed with the occurrence data points and AOO was measured as the sum of occupied cells in a uniform grid scaled to 4 km² (2 x 2 km) cell size. The authors, recognizing that these values are probably underestimated, proposed upper estimates for EOO and AOO based on a species distribution model (SDM). The MCP around the outermost limits of the SDM area was the upper EOO (EOOup), and the SDM area within EOO was the upper AOO (AOOup).

    The estimated values from Freitas et al. (2019) were:
    EOO = 24,555.85 km²; EOOup = 30,697.58 km²;
    AOO = 228.0 km²; AOOup = 1,827.39 km².

    Both EOO does not meet the threshold for Vulnerable (EOO < 20,000 km²), but the estimation based on the occurrence data points closely approaches it.

    Both AOO meet the threshold for the Vulnerable category (< 2,000 km²), being the lower estimation within the ‘Endangered’ category (AOO < 500 km²).

    Reference cited:
    Freitas, G.H.S., Costa, L.M., Chaves, A. V., Vasconcelos, M.F., Ribeiro, L.C., Silva, J.C., Souza, R.A., Santos, F.R. & Rodrigues, M. (2019a). Geographic range and conservation of the Cipo Canastero Asthenes luizae, an endemic furnariid of Brazilian sky islands. Bird Conservation International 1–16.

  2. Lílian M. Costa & Guilherme H. S. Freitas says:

    2) Severely fragmented distribution/population:

    The species’ population is probably severely fragmented. Its distribution is naturally fragmented in particular habitats at the mountaintops and this fragmentation is probably increasing with the habitat changes by human causes (see Fernandes et al. 2018, Monteiro et al. 2018).

    Although the Diamantina Plateau is the larger ‘sky island’ mapped by Freitas et al. (2019a), holding more than half (55%) of the total area of all ‘sky islands’, it is not fully occupied (only 13% of it overlaps with the SDM; Freitas et al. 2019a).
    When we calculated how much of each ‘sky island’ is intersected with the area of the SDM (data from Table 1 in Freitas et al. 2019a), we found that none holds more than one third of the total SDM area.

    Additionally, the distribution within the sky islands is also very fragmented.
    Cipo Canastero inhabits only one microhabitat (rock outcrops) of the habitat types occurring in campos rupestres ecosystem, and not all rock outcrops are suitable because their home-ranges are settled in the larger patches along ridges (Freitas et al. 2019b; see also Figure S1 in Freitas et al. 2019a).

    Also, there is evidence that the species has poor dispersal ability (preliminary data suggest natal dispersal distances of only ~0.7–1.1 km; Freitas et al. 2019b), and that recent gene flow among the isolated populations are low or nonexistent (preliminary genetic study found haplotypes exclusive to some populations; Xavier et al. 2013).

    So, we might infer that most (>50%) of the total area of occupancy is in habitat patches that are small and isolated.

    Reference cited:

    Fernandes, G.W., N.P.U. Barbosa, B. Alberton, A. Barbieri, R. Dirzo, F. Goulart, T.J. Guerra, L.P.C. Morellato, & R.R.C. Solar (2018) The deadly route to collapse and the uncertain fate of Brazilian rupestrian grasslands. Biodiversity and Conservation 27: 2587–2603.

    Freitas, G.H.S., Costa, L.M., Chaves, A. V., Vasconcelos, M.F., Ribeiro, L.C., Silva, J.C., Souza, R.A., Santos, F.R. & Rodrigues, M. (2019a) Geographic range and conservation of the Cipo Canastero Asthenes luizae, an endemic furnariid of Brazilian sky islands. Bird Conservation International 1–16. DOI: https://doi.org/10.1017/S0959270919000418.

    Freitas, G.H.S., Costa, L.M. & Rodrigues, M. (2019b) Home-ranges, population densities, vocal behavior, and post-fledging movements of Cipo Canasteros (Asthenes luizae, Furnariidae), a rock-specialist endemic of the highlands of eastern Brazil. J. F. Ornithol. 90: 215–228. DOI: https://doi.org/10.1111/jofo.12308.

    Monteiro, L., N. Machado, E. Martins, N. Pougy, M. Verdi, G. Martinelli & R. Loyola (2018) Conservation priorities for the threatened flora of mountaintop grasslands in Brazil. Flora 238: 234–243.

    Xavier, L.S.S., Chaves, A.V., Freitas, G.H.S., Costa, L.M., Vasconcelos, M.F., Rodrigues, M. & Santos, F.R. (2013) Diversidade genética de Asthenes luizae (Aves: Furnariidae) na Serra do Cipó e suas relações filogeográficas na Cadeia do Espinhaço, MG. In: XXI Semana de Iniciação Científica da UFMG – Conhecimento e Cultura. Belo Horizonte: Universidade Federal de Minas Gerais.

  3. Lílian M. Costa & Guilherme H. S. Freitas says:

    3) Global population size:

    The projection of Costa (2015) mentioned in the forum post was based on an earlier estimate of the area of suitable habitat, that should be replaced for the more refined and updated AOO estimates from Freitas et al. (2019a).

    By doing this and keeping the densities previously estimated by Costa (2015) we found the following average numbers of paired individuals: 1,596 (range = 456 – 3,648) when based in AOO, and 12,792 (range = 3,655 – 29,238) when based in AOOup.

    These mean estimates would qualify the species as Endangered (<2,500 mature individuals) or possibly as Near-Threatened by approaching the Vulnerable threshold (<10,000 mature individuals).

    Reference cited:

    Costa, L. M. (2015) História natural, demografia, viabilidade populacional e conservação de Asthenes luizae (Furnariidae), ave endêmica dos campos rupestres da Cadeia do Espinhaço, Minas Gerais. Belo Horizonte: Universidade Federal de Minas Gerais.

    Freitas, G.H.S., Costa, L.M., Chaves, A. V., Vasconcelos, M.F., Ribeiro, L.C., Silva, J.C., Souza, R.A., Santos, F.R. & Rodrigues, M. (2019a) Geographic range and conservation of the Cipo Canastero Asthenes luizae, an endemic furnariid of Brazilian sky islands. Bird Conservation International 1–16. DOI: https://doi.org/10.1017/S0959270919000418.

  4. Lílian M. Costa & Guilherme H. S. Freitas says:

    4) The threat of fire:

    It is important to add that most of the fire occurrences in the species range are not as unusual as suggested by the statement that they came from ‘tourist campfires’ or ‘escaped agricultural fires’. Instead, they are frequent and mainly intentional, mostly to renewing the grassland vegetation for use as natural pastures (Ribeiro & Figueira 2011, Alvarado et al. 2014). There are also criminal fires derived from conflicts between people and protected areas (Figueira et al. 2016, Monteiro et al. 2018).
    Although the campos rupestres is a fire-prone ecosystem, human changes in the fire regime increases fire frequency and intensity, preventing adequate recovering of the ecosystem and threatening the campos rupestres biodiversity (Figueira et al. 2016).

    There are some analyzed data about fire in the Serra do Cipó region, including Serra do Cipó National Park and the Morro da Pedreira Environmental Protection Area (covering part of two of the ‘sky islands’ from Freitas et al. 2019a – Cipó South and Cipó North).
    Over a 30-year period, almost half (48%) of this area was burned every 1-2 years (Alvarado et al. 2017). Some areas neighboring the National Park burn more than once within a year (Ribeiro & Figueira 2011).
    Fire events are mainly observed throughout the dry season, with a peak in September-October (Alvarado et al. 2014), coinciding with the peak of the Cipo Canastero’s breeding activity (Costa 2015).

    The predominant vegetation of the mountaintops, composed mostly of grasses, that became dehydrated in the dry season, combined with the strong winds of the high altitudes, predispose the campos rupestres to extensive fires (Ribeiro & Figueira 2011).
    A single fire event can spread over large areas. In some years, the burned area was larger than 300 km² (Alvarado et al. 2017), which is larger than the SDM area within most of the ‘sky islands’.
    However, the number of locations based on the threat of fire would still likely be greater than ten.

    5) Other threats:

    One additional threat to the species is mining. Almost half (47%) of the species’ modeled suitable habitat (by Pena et al. 2017) might be currently either directly or indirectly affected by mining activities (Pena et al. 2017).

    Another threat to be considered for management is the cattle ranching. Besides the potential direct damage on soil and plant structure, altering the habitat, it is related to the occurrence of invasive grasses, fire, and Molothrus spp. (Kolbek & Alves 2008). As pointed before, Shiny Cowbird (Molothrus bonariensis) is an important threat, and can reduce the reproductive success of a female Cipo Canastero by 50% (Costa 2015).

    For a more complete list of threats to the campos rupestres see Fernandes et al. (2018) and Monteiro et al. (2018).

    References cited:

    Alvarado, S.T., Fornazari, T., Cóstola, A., Morellato, L.P.C. & Silva, T.S.F. (2017) Drivers of fire occurrence in a mountainous Brazilian cerrado savanna: Tracking long-term fire regimes using remote sensing. Ecological Indicators 78: 270-281.

    Costa, L. M. (2015) História natural, demografia, viabilidade populacional e conservação de Asthenes luizae (Furnariidae), ave endêmica dos campos rupestres da Cadeia do Espinhaço, Minas Gerais. Belo Horizonte: Universidade Federal de Minas Gerais.

    Fernandes, G.W., N.P.U. Barbosa, B. Alberton, A. Barbieri, R. Dirzo, F. Goulart, T.J. Guerra, L.P.C. Morellato, & R.R.C. Solar (2018) The deadly route to collapse and the uncertain fate of Brazilian rupestrian grasslands. Biodiversity and Conservation 27: 2587–2603.

    Figueira, J.E.C., K.T. Ribeiro, M.C. Ribeiro, C.M. Jacobi, H. França, A.C.O. Neves, A.A. Conceição, F.A. Mourão, J.M. Souza & C.A.K. Miranda (2016) Fire in Rupestrian Grasslands: Plant Response and Management. In: Ecology and Conservation of Mountaintop Grasslands in Brazil. pp. 415–448.

    Freitas, G.H.S., Costa, L.M., Chaves, A. V., Vasconcelos, M.F., Ribeiro, L.C., Silva, J.C., Souza, R.A., Santos, F.R. & Rodrigues, M. (2019a) Geographic range and conservation of the Cipo Canastero Asthenes luizae, an endemic furnariid of Brazilian sky islands. Bird Conservation International 1–16. DOI: https://doi.org/10.1017/S0959270919000418.

    Kolbek, J. & Alves, R.J.V. (2008) Impacts of cattle, fire and wind in rocky savannas, southeastern Brazil. Acta Univ. Carol. Environ. 22: 111–130.

    Monteiro, L., N. Machado, E. Martins, N. Pougy, M. Verdi, G. Martinelli & R. Loyola (2018) Conservation priorities for the threatened flora of mountaintop grasslands in Brazil. Flora 238: 234–243.

    Pena, J.C.C., Goulart, F., Fernandes, G.W., Hoffmann, D., Leite, F.S., Santos, N.B., … & Rodrigues, M. (2017) Impacts of mining activities on the potential geographic distribution of eastern Brazil mountaintop endemic species. Perspectives in ecology and conservation, 15: 172-178.

    Ribeiro, M.C., & Figueira, J.E.C. (2011) Uma abordagem histórica do fogo no Parque Nacional da Serra do Cipó, Minas Gerais–Brasil. Biodiversidade Brasileira, (2), 212-227.

  5. Red List Team (BirdLife International) says:

    Many thanks to everyone who has contributed to this discussion. We greatly appreciate the time and effort invested by so many people in commenting. The window for consultation is now closed. We will analyse and interpret the new information and post a preliminary decision on this species’s Red List status on this page in early July.

    Thank you once again,
    BirdLife Red List Team

  6. Red List Team (BirdLife International) says:

    Preliminary proposal

    Thank you very much for the detailed and extremely helpful comments on this proposal, and for providing the GIS file of the distribution model. We have analysed and interpreted this additional information as follows:

    Criterion B: By creating a minimum convex polygon around the polygons of the species distribution model (SDM; as received from L. M. Costa via email), and finding the geodesic area, we calculate the extent of occurrence (EOO) to be 31,300 sqkm. The small difference in the EOO value provided by Lílian M. Costa & Guilherme H. S. Freitas above may be the result of a difference in the method used to calculate area, or in the map projection used. In both cases, the value does not approach the threshold for Vulnerable under Criterion B1 (20,000 sqkm), so we retain our assessment of Least Concern under this criterion. We have not created a minimum convex polygon around the species’s point records alone, as we consider this likely to produce an underestimate. Based on the area of 4 sqkm grid cells intersecting the polygons of the SDM, we estimate the AOO to be 4,736 sqkm. By removing some area of the modelled range polygons without nearby records, we produced a more conservative AOO estimate of 3,672 sqkm. These values do not approach the threshold for Vulnerable under Criterion B2.
    It is not valid to use the area of the modelled range polygons as the AOO, because the AOO should be estimated at the scale of 4 sqkm cells, to ensure consistency of comparison against the thresholds for Criterion B2 (IUCN Standards and Petitions Committee 2019). Although the minimum AOO provided above falls beneath the threshold for Endangered under Criterion B2, we consider that an AOO produced only based on point localities is likely to be an underestimate, and that including the modelled habitat provides a more realistic estimate of occupied range. Taking into account the uncertainty around the AOO, the species could qualify as Near Threatened, or Least Concern.
    According to the IUCN Red List Guidelines (IUCN Standards and Petitions Committee 2019), a species may be classified as severely fragmented when >50% of its total area of occupancy is in habitat patches that are smaller than would be required to support a viable population and separated from other habitat patches by a large distance. “For example, for many vertebrates, subpopulations of fewer than 100 individuals may be considered too small to be viable.” (IUCN Standards and Petitions Committee 2019). By multiplying the area of each ‘sky island’ by the proportion of each sky island that falls within the species distribution model (from Table 1 in Freitas et al. 2019) to find the area of the distribution model in each, then dividing the minimum population estimate (3,034 mature individuals – see Criterion C below) between the sky islands according to their relative area of SDM, we estimate that three sky islands (Serra da Formosa, Itacambira and Serra do Ambrosio) are likely to hold fewer than 100 mature individuals each, with a combined total of 142 mature individuals. This is based on the minimum population estimate, and only considering mature individuals. Therefore, we do not consider it likely that >50% of the species’s area of occupancy is in patches that are smaller than would be required to support a viable population, and thus the species is not severely fragmented.
    Following analysis of the further information provided about the threat of fires, we acknowledge the severity of the threat, but it still appears that more than ten fires would need to occur before the species’s entire range was destroyed or severely reduced, meaning that there are more than ten locations. Given that the threshold for Criterion B2 is probably not met, and only one condition (a continuing decline) is met, our preliminary assessment under this criterion is Least Concern.

    Criterion C: By multiplying the density estimates provided by Costa (2015) by their AOO estimates (as provided in their comment), L. M. Costa and G. H. S. Freitas have provided revised population size estimates. Two new population estimates have been provided; one is based on an AOO derived from 4 sqkm squares overlaying records, and the other is based on the area of a species distribution model. We do not consider it valid to derive the population estimate from the area of the 4 sqkm grid squares because this does not represent the area of habitat (but is used under Criterion B2 to standardise area measurements for application of the thresholds). We consider that the estimate based on the AOO is more likely to be accurate: 12,792 (3,655-29,238) paired individuals.
    By removing some area of the modelled range polygons without nearby records (as under Criterion B above), we produced a more conservative version of the modelled range, which represented 83% of the original area of the SDM. Multiplying the above population estimated by 83% produced a more conservative population estimate of 10,617 (3,034-24,268). Based on the estimates previously provided by Costa (2015), and those detailed here, the population size is estimated to fall within the range 3,000-48,000 mature individuals (here rounded), with a best estimate of 10,600. The lower end of the range falls beneath the threshold for Vulnerable and the best estimate closely approaches the threshold. As described in the original post, for a species to qualify as threatened under Criterion C, other conditions must also be met. There is an inferred continuing decline (Costa 2015), but the rate of decline has not been estimated, so Criterion C1 cannot be used. There is more than one subpopulation and Criterion C2a(ii) does not apply, and there are no extreme fluctuations, so Criterion C2b is not relevant.
    The range is divided into three areas separated by some distance and there are at least three subpopulations. However, it is likely that there are further subpopulation divisions within these blocks, particularly the southern block, due to narrower gaps of around 10 km between areas of suitable elevation. A study of the species’s habitat and elevational range identified nine ‘sky islands’ within the range, which may represent separate subpopulations (Freitas et al. 2019). By multiplying the area of each ‘sky island’ by the proportion of each sky island that falls within the SDM (from Table 1 in Freitas et al. 2019) to find the area of the distribution model in each, then dividing the population estimate (3,000 – 48,000 mature individuals; Costa 2015) between the sky islands according to their relative area of SDM, we estimate that the largest sky island (Cipó North) is likely to hold 974-15,287 mature individuals, with a best estimate of 3,404. The lower end of this range would meet the Vulnerable threshold for Criterion C2a(i), but the best estimate is significantly larger than this threshold. Although it cannot be ruled out that the species could meet the thresholds for population size and number of mature individuals in each subpopulation to qualify for Vulnerable under Criterion C2a(i), we consider that the true population and subpopulation sizes are more likely to exceed these thresholds, so our preliminary assessment under this criterion is Near Threatened.

    Threats: Following receipt of comments about threats, we have updated our text about threats in the species’s assessment, which will be published in December 2020 or January 2021.

    Based on the information presented here, our preliminary proposal for the 2020 Red List would be to list Cipo Canastero as Near Threatened, approaching the thresholds for Criterion C2a(i).

    There is now a period for further comments until the final deadline in mid-July, after which the recommended categorisations will be put forward to IUCN.

    Please note that we will then only post final recommended categorisations on forum discussions where these differ from the initial proposal.

    The final 2020 Red List categories will be published on the BirdLife and IUCN websites in December 2020 or January 2021, following further checking of information relevant to the assessments by both BirdLife and IUCN.

Leave a Reply

You have to agree to the comment policy.

All comments must follow the rules of usage.

This site uses Akismet to reduce spam. Learn how your comment data is processed.