Limits...
The Butterflies of Barro Colorado Island, Panama: Local Extinction since the 1930s.

Basset Y, Barrios H, Segar S, Srygley RB, Aiello A, Warren AD, Delgado F, Coronado J, Lezcano J, Arizala S, Rivera M, Perez F, Bobadilla R, Lopez Y, Ramirez JA - PLoS ONE (2015)

Bottom Line: When confirmed, local extinction was often due to the loss of host-plant species.Coupled with low dispersal power, this reduced availability of host plants has probably caused the local extinction of some butterfly species.Many more bird than butterfly species have been lost from BCI recently, confirming that small preserves may be far more effective at conserving invertebrates than vertebrates and, therefore, should not necessarily be neglected from a conservation viewpoint.

View Article: PubMed Central - PubMed

Affiliation: Smithsonian Tropical Research Institute, Apartado 0843-03092, Panama City, Republic of Panama; Faculty of Science, University of South Bohemia and Institute of Entomology, Biology Centre of Czech Academy of Sciences, Branišovská 31, 370 05, České Budějovice, Czech Republic; Universidad de Panamá, Maestria de Entomologia, 080814, Panama City, Republic of Panama.

ABSTRACT
Few data are available about the regional or local extinction of tropical butterfly species. When confirmed, local extinction was often due to the loss of host-plant species. We used published lists and recent monitoring programs to evaluate changes in butterfly composition on Barro Colorado Island (BCI, Panama) between an old (1923-1943) and a recent (1993-2013) period. Although 601 butterfly species have been recorded from BCI during the 1923-2013 period, we estimate that 390 species are currently breeding on the island, including 34 cryptic species, currently only known by their DNA Barcode Index Number. Twenty-three butterfly species that were considered abundant during the old period could not be collected during the recent period, despite a much higher sampling effort in recent times. We consider these species locally extinct from BCI and they conservatively represent 6% of the estimated local pool of resident species. Extinct species represent distant phylogenetic branches and several families. The butterfly traits most likely to influence the probability of extinction were host growth form, wing size and host specificity, independently of the phylogenetic relationships among butterfly species. On BCI, most likely candidates for extinction were small hesperiids feeding on herbs (35% of extinct species). However, contrary to our working hypothesis, extinction of these species on BCI cannot be attributed to loss of host plants. In most cases these host plants remain extant, but they probably subsist at lower or more fragmented densities. Coupled with low dispersal power, this reduced availability of host plants has probably caused the local extinction of some butterfly species. Many more bird than butterfly species have been lost from BCI recently, confirming that small preserves may be far more effective at conserving invertebrates than vertebrates and, therefore, should not necessarily be neglected from a conservation viewpoint.

No MeSH data available.


Related in: MedlinePlus

Plot of the scores of sampling years in axes 1 and 2 of the NMDS.Years are linked chronologically by a solid line. Pie charts indicate for each year the proportion of abundance accounted by (in clockwise order) Hesperiidae (black), Lycaenidae (white), Nymphalidae (grey), Papilionidae (black stippled), Pieridae (white stippled) and Riodinidae (grey squared).
© Copyright Policy
Related In: Results  -  Collection

License
getmorefigures.php?uid=PMC4549329&req=5

pone.0136623.g002: Plot of the scores of sampling years in axes 1 and 2 of the NMDS.Years are linked chronologically by a solid line. Pie charts indicate for each year the proportion of abundance accounted by (in clockwise order) Hesperiidae (black), Lycaenidae (white), Nymphalidae (grey), Papilionidae (black stippled), Pieridae (white stippled) and Riodinidae (grey squared).

Mentions: The NDMS analysis clearly separated recent monitoring years (2008–2013) from older years, including 1932, along Axis 1 of the ordination (Fig 2). Multiple regressions indicated that only the variable Year itself explained significantly the scores of years on Axis 1 (F = 32.02, p<0.001, R2 = 0.775, n = 10), whereas sampling effort (no. of individuals recorded) explained to a lesser extent the formation of Axis 2 (F = 5.61, p = 0.045, R2 = 0.339, n = 10). The proportion of individuals recorded per family also varied along years, with, for example, a higher proportion of Pieridae (especially Itaballia spp.) and Riodinidae (Detritivora spp.) and a lower proportion of Hesperiidae being recorded during more recent years, as compared with Year 1932 (Fig 2). This was confirmed by a Chi-square test comparing the number of individuals recorded in each family for the old vs. recent period (Chi-square = 797.3, p < 0.0001, d.f. = 5). The Hesperiidae decreased from 36% to 10% of records, while Pieridae and Riodinidae rose from 6% to 20% and from 12% to 24%, respectively. In contrast, the number of species per family remained similar between the old vs. recent period (Chi-square = 5.3, p = 0.374, d.f. = 5). There was also a negative correlation between the number of individuals per species recorded in years 1932 and 2013 (rs = -0.301, p<0.001, n = 327) and, similarly, between the number of individuals per species recorded during the old and recent periods (rs = -0.252, p<0.001, n = 529).


The Butterflies of Barro Colorado Island, Panama: Local Extinction since the 1930s.

Basset Y, Barrios H, Segar S, Srygley RB, Aiello A, Warren AD, Delgado F, Coronado J, Lezcano J, Arizala S, Rivera M, Perez F, Bobadilla R, Lopez Y, Ramirez JA - PLoS ONE (2015)

Plot of the scores of sampling years in axes 1 and 2 of the NMDS.Years are linked chronologically by a solid line. Pie charts indicate for each year the proportion of abundance accounted by (in clockwise order) Hesperiidae (black), Lycaenidae (white), Nymphalidae (grey), Papilionidae (black stippled), Pieridae (white stippled) and Riodinidae (grey squared).
© Copyright Policy
Related In: Results  -  Collection

License
Show All Figures
getmorefigures.php?uid=PMC4549329&req=5

pone.0136623.g002: Plot of the scores of sampling years in axes 1 and 2 of the NMDS.Years are linked chronologically by a solid line. Pie charts indicate for each year the proportion of abundance accounted by (in clockwise order) Hesperiidae (black), Lycaenidae (white), Nymphalidae (grey), Papilionidae (black stippled), Pieridae (white stippled) and Riodinidae (grey squared).
Mentions: The NDMS analysis clearly separated recent monitoring years (2008–2013) from older years, including 1932, along Axis 1 of the ordination (Fig 2). Multiple regressions indicated that only the variable Year itself explained significantly the scores of years on Axis 1 (F = 32.02, p<0.001, R2 = 0.775, n = 10), whereas sampling effort (no. of individuals recorded) explained to a lesser extent the formation of Axis 2 (F = 5.61, p = 0.045, R2 = 0.339, n = 10). The proportion of individuals recorded per family also varied along years, with, for example, a higher proportion of Pieridae (especially Itaballia spp.) and Riodinidae (Detritivora spp.) and a lower proportion of Hesperiidae being recorded during more recent years, as compared with Year 1932 (Fig 2). This was confirmed by a Chi-square test comparing the number of individuals recorded in each family for the old vs. recent period (Chi-square = 797.3, p < 0.0001, d.f. = 5). The Hesperiidae decreased from 36% to 10% of records, while Pieridae and Riodinidae rose from 6% to 20% and from 12% to 24%, respectively. In contrast, the number of species per family remained similar between the old vs. recent period (Chi-square = 5.3, p = 0.374, d.f. = 5). There was also a negative correlation between the number of individuals per species recorded in years 1932 and 2013 (rs = -0.301, p<0.001, n = 327) and, similarly, between the number of individuals per species recorded during the old and recent periods (rs = -0.252, p<0.001, n = 529).

Bottom Line: When confirmed, local extinction was often due to the loss of host-plant species.Coupled with low dispersal power, this reduced availability of host plants has probably caused the local extinction of some butterfly species.Many more bird than butterfly species have been lost from BCI recently, confirming that small preserves may be far more effective at conserving invertebrates than vertebrates and, therefore, should not necessarily be neglected from a conservation viewpoint.

View Article: PubMed Central - PubMed

Affiliation: Smithsonian Tropical Research Institute, Apartado 0843-03092, Panama City, Republic of Panama; Faculty of Science, University of South Bohemia and Institute of Entomology, Biology Centre of Czech Academy of Sciences, Branišovská 31, 370 05, České Budějovice, Czech Republic; Universidad de Panamá, Maestria de Entomologia, 080814, Panama City, Republic of Panama.

ABSTRACT
Few data are available about the regional or local extinction of tropical butterfly species. When confirmed, local extinction was often due to the loss of host-plant species. We used published lists and recent monitoring programs to evaluate changes in butterfly composition on Barro Colorado Island (BCI, Panama) between an old (1923-1943) and a recent (1993-2013) period. Although 601 butterfly species have been recorded from BCI during the 1923-2013 period, we estimate that 390 species are currently breeding on the island, including 34 cryptic species, currently only known by their DNA Barcode Index Number. Twenty-three butterfly species that were considered abundant during the old period could not be collected during the recent period, despite a much higher sampling effort in recent times. We consider these species locally extinct from BCI and they conservatively represent 6% of the estimated local pool of resident species. Extinct species represent distant phylogenetic branches and several families. The butterfly traits most likely to influence the probability of extinction were host growth form, wing size and host specificity, independently of the phylogenetic relationships among butterfly species. On BCI, most likely candidates for extinction were small hesperiids feeding on herbs (35% of extinct species). However, contrary to our working hypothesis, extinction of these species on BCI cannot be attributed to loss of host plants. In most cases these host plants remain extant, but they probably subsist at lower or more fragmented densities. Coupled with low dispersal power, this reduced availability of host plants has probably caused the local extinction of some butterfly species. Many more bird than butterfly species have been lost from BCI recently, confirming that small preserves may be far more effective at conserving invertebrates than vertebrates and, therefore, should not necessarily be neglected from a conservation viewpoint.

No MeSH data available.


Related in: MedlinePlus