Current warming and likely future impacts - Climate Change: Biological and Human Aspects - page 291

data for all 1902 species were taken between 2001 and 2007 from the Barva Transect,
a continuously forested corridor ascending 2900 m up an elevation gradient from
La Selva Biological Station, near sea level, to the top of Volcan Barva, in Costa
Rica. To explore the question of tropical lowland ecological attrition we must ask
whether tropical lowland species are already living near the thermal optimum of
their climatic niche, above which fitness would decline in the absence of acclimation
or adaptation. For plants, especially, because temperature and precipitation interact
strongly through transpiration water loss, the answer includes precipitation change
as well that of temperature. With climate change, as some parts of the tropics will
see more rainfall, and others less, the picture becomes less clear. Current evidence,
however, suggests that contemporary warming of around 0.25 ◦ C decade − 1 since
1975 in the tropical lowlands has already driven global mean temperature to within
approximately 1 ◦ C of the Earth's maximum temperature over the past couple of
million years. The Cowell-Brehm team used a simple graphical model that relied on
temperature to assess potential altitude range shifts for their fauna and flora transect
data. Assuming an IPCC 2007 warming of 3.5 ◦ C by the year 2100, a 600 m vertical
compensatory shift in their Barva Transect species' range would be required. About
half (53%) of the 1902 species in their study (those that are currently in the lowest
elevations) are candidates for lowland biotic attrition, and about half (51%) may be
faced with range-shift gaps. The potential for mountains-top extinctions for these
groups on the Barva Transect is minimal for a 600 m shift in isotherms but begins to
be more significant at about a 1000 m range shift (which could occur by the early to
mid-22nd century if the IPCC 2007 longer-term scenario forecasts are correct). So,
although mountain-top extinction has received quite a bit of recent attention in the
ecological literature, this study suggests that in the near term a far greater proportion
of the tropical species are threatened with lowland attrition or are challenged by
landscape fragmentation (natural and anthropogenic). Many face more than one
challenge.
A broadly similar result came in 2010 from Kenneth Feeley and Miles Silman,
at Wake Forest University, Northern Carolina. They looked at more than 2000 plant
species from tropical South America and their thermal niches to see how they might
respond to 1-5 ◦ C of warming. Looking at the results on the basis of observed
thermal niches, they predicted an almost complete loss of plant diversity in most
South American tropical forests due to 5 ◦ C warming, but correcting for possible
niche truncation they estimated that most forests will retain 50-70%, or more, of
their current species richness. This Feeley-Silman result, and the Cowell-Brehm
et al. result, both very broadly suggest that lowland tropical areas in the warmest part
of the Amazon might see an attrition of roughly between a third and half of species
with warming of 3.5-5 ◦ C. This is the temperature rise that the central Amazon is
likely to experience by the end of the 21st century under the 2007 IPCC A2 B-a-U
emissions scenario. (Of course the above does not include precipitation change, hence
forecast summer drought, in some parts of a warmer Amazonia, which would provide
additional plant stress.)
Finally, before leaving the subject of ecological and species response to climate
change in the tropics it is worth recalling some of the ground previously covered
when looking at the last glacial (section 4.5). Because tropical lowlands were refugia