by Brian Tomasik
First written: 2008; major additions: 2013; last update: 15 Jan. 2017

## Summary

Human environmental choices have vast implications for wild animals, and one of our largest ecological impacts is climate change. Each human in the industrialized world may create or prevent in a potentially predictable way at least millions of insects and potentially more zooplankton per year by his or her greenhouse-gas emissions. Is this influence net good or net bad? This question is very complicated to answer and takes us from examinations of tropical-climate expansion, sea ice, and plant productivity to desertification, coral reefs, and oceanic-temperature dynamics. On balance, I'm extremely uncertain about the net impact of climate change on wild-animal suffering; my probabilities are basically 50% net good vs. 50% net bad when just considering short-term animal suffering. Since other people care a lot about preventing climate change, and since climate change might destabilize prospects for a cooperative future, I currently think it's best to err on the side of reducing our greenhouse-gas emissions where feasible, but my low level of confidence reduces my fervor about the issue in either direction. That said, I am fairly confident that biomass-based carbon offsets, such as rainforest preservation, are net harmful for wild animals.

## Introduction

Life in the wild contains vast amounts of suffering, and while it's sometimes assumed that humans are powerless to do anything about it, this is in fact untrue: Every day each of us makes ecological decisions that have major implications for how many wild animals exist on Earth and what kinds of reproductive strategies they pursue. A challenging but important task is to determine the sign of the net impact of various environmental policies on how many wild animals exist, and hence on the extent of wild-animal suffering.

This piece outlines some considerations along those lines for climate change. The potential implications are wide-ranging, because so many of our day-to-day environmental choices affect greenhouse-gas emissions: driving, meat consumption, electricity and hot-water use, whether to recycle cans, and many more.

## How much could climate change matter?

Here's a sample back-of-the-envelope calculation:

• In 2012, the world emitted 34.5 billion tons of CO2, and a person in the USA contributed on average 16.4 tons of CO2.
• Suppose, naively, that one year of CO2 emissions contributes to one extra year of the effects of climate change. In fact, this is probably very conservative, since climate change may have prolonged repercussions even if CO2 emissions end. Indeed, absent geoengineering, some CO2 emissions will continue to affect radiative forcing for millennia.
• Also assume that the global-warming contribution of each ton of CO2 is the same. In practice, the marginal effect of an additional ton of CO2 might increase or decrease depending on whether the warming exhibits increasing or decreasing returns to scale, respectively.
• Take the world population of mammals to be 1012 and of insects to be 1018.
• Let's say we could, after deeper analysis of the type described in this essay, predict that climate change would increase or decrease wild-animal populations by a net 2% in expectation. (This estimate can contain uncertainty -- e.g., a 50% chance of increasing by 7% and a 50% chance of decreasing by 3%.)

Then each person in the USA would, via the ecological effects of CO2 emissions alone, create or prevent in one year
$\frac{16.4\:\textrm{tons}}{34.5\cdot10^9\:\text{tons}}\:\cdot2\:\text{percent}\cdot\:10^{12}\:\text{mammals}\:\cdot\:1\:\text{assumed}\:\text{year}\:\text{of}\:\text{impact}=\\\\10\:\text{mammal-years}$
and
$\frac{16.4\:\textrm{tons}}{34.5\cdot10^9\:\text{tons}}\:\cdot2\:\text{percent}\cdot\:10^{18}\:\text{insects}\:\cdot\:1\:\text{assumed}\:\text{year}\:\text{of}\:\text{impact}=\\\\10\:\text{million}\:\text{insect-years}$
Note that since insects have vast numbers of offspring per parent pair and may often live less than a year, an insect-year actually means tens to hundreds of post-hatching infant insect deaths. Thus, one person's global-warming impact might create or prevent net billions of insect deaths per year. Clearly a lot is at stake with this question.

Because most wild animals endure more suffering than happiness in the wild (especially short-lived ones like insects), we would hope to decrease wild-animal populations via our actions.

## Ways climate change might increase animal populations

There is no shortage of information on specific effects of climate change on various ecosystems, species, and biome patterns; e.g., Chapter 4 of the IPCC Fourth Assessment Report is a great place to start. Below I examine a few important factors, but there are many more I've left out.

### Effects on land

#### More insects

Several studies predict increases in insect populations due to climate change.

This textbook explains that "The western corn rootworm and the Colorado potato beetle are moving further and further into northern Europe [...], while remote areas in Alaska are experiencing range expansions of agricultural and forest pests."

From Erik E. Stange and Matthew P. Ayres, Climate Change Impacts: Insects:

• Warmer temperatures associated with climate changes will tend to influence (and frequently amplify) insect species' population dynamics directly through effects on survival, generation time, fecundity and dispersal. [...]
• Insect populations in mid- to high latitudes are expected to benefit most from climate change through more rapid development and increased survival. Much less is known about the effects of increased warming on tropical insect species.
• Insect species' mortality may decrease with warmer winter temperatures, thereby leading to poleward range expansions.

Note that reduced mortality over winter may mean reduced r-selection and lower death rates, so this is a silver lining to the increase in total insect populations in higher latitudes.

From Global Warming Could Trigger Insect Population Boom: [Melanie Frazier] and her colleagues looked at 65 insect species and found a correlation between warm climes and population growth across the board [...]. See also

In polar regions that are expected to experience among the most rapid impact from climate warming, springtails have shown contrasting responses to warming in experimental warming studies.[72] There are negative,[73][74] positive[75][76] and neutral responses reported.[74][77] Neutral responses to experimental warming have also been reported in studies of non-polar regions.[78]

This study found mixed effects from warming on bug populations. Based on the study's abstract, here's a summary of responses to warming:

 Land type Taxa that increased in abundance Taxa that decreased in abundance glade Oribatida Collembola, Gamasida fellfield Oribatida Collembola, Gamasida heath "no strong effects of warming" "no strong effects of warming"

The authors conclude that "the observed changes are probably linked to changes in food availability more than direct climatic influences."

Insects in warmer climates can also have more generations per year, which means more deaths per year. Here's one example comparison of temperate vs. tropical bugs:

Mulberry silkworms can be categorized into three different but connected groups or types. The major groups of silkworms fall under the univoltine ('uni-'=one, 'voltine'=brood frequency) and bivoltine categories. The univoltine breed is generally linked with the geographical area within greater Europe. The eggs of this type hibernate during winter due to the cold climate, and cross-fertilize only by spring, generating silk only once annually. The second type is called bivoltine and is normally found in China, Japan, and Korea.

The breeding process of this type takes place twice annually, a feat made possible through the slightly warmer climates and the resulting two lifecycles. The polyvoltine type of mulberry silkworm can only be located in the tropics. The eggs are laid by female moths and hatch within nine to 12 days, so the resulting type can have up to eight separate lifecycles throughout the year.[5]

#### Pathogens and parasites

Diseases and parasitic organisms tend to be more common in tropical regions of the Earth. Presumably global warming would spread their abundances to higher latitudes. This could make animals' lives worse due to higher mortality and morbidity?

On the other hand, if diseases and parasites have enough of a top-down-control effect to reduce total animal abundance across all species, then they could also reduce suffering?

### Effects in the ocean

#### Elimination of sea ice

Over the last 30 years, however, the Arctic has warmed, and larger areas of the Arctic Ocean are now free of sea ice in the summer, which means phytoplankton are getting more sunlight. The result is that phytoplankton productivity has increased by about 20 percent based on satellite estimates of the amount of chlorophyll in the water.

#### Geoengineering

As climate change gets worse, there will be more pressure for geoengineering to combat CO2. This may include ocean fertilization, which would increase phytoplankton and probably populations of at least some of its consumers: "Adding urea to the ocean can cause blooms of phytoplankton that is source of food of zooplankton and in turn feed for fish. This is in turn expected to increase the sustainable fish catch.[29]"

This textbook includes the following figure showing the (alarming) results of an early experiment with iron fertilization:

the biosphere considered as a whole has managed to expand the amount of solar energy captured for metabolism to around 5% [...], limited by the nonuniform presence of key nutrients across the Earth’s surface -- primarily fresh water, phosphorus, and nitrogen. Life on Earth is not free-energy-limited because, up until recently, it has not had the intelligence and mega-engineering to distribute Earth’s resources to all of the places solar energy happens to fall, and so it is, in most places, nutrient-limited [...].

This suggests a scary possibility that if humans tried, they might be able to amplify Earth's biological energy utilization appreciably higher.

Less speculatively, climate change creates more pressure to preserve rainforests and other biomass-rich habitats, which allows for the continuation of vast amounts of wild-animal suffering.

## Ways climate change might decrease animal populations

### In general, it takes time to adapt

We might expect populations to decrease at least in the short term because animals have evolved to squeeze out maximal populations in their current environments. It will take time for those species that don't go extinct to adapt to the changes.

From Biology by Robert J. Brooker, Eric P. Widmaier, Linda E. Graham, and Peter D. Stiling, pp. 1141-42:

Assuming this scenario of gradual global warming is accurate, we need to consider what the consequences might be for plant and animal life. At the beginning of the chapter, we saw how global warming is believed to be contributing to the decline and extinction of some amphibian species. Although many species can adapt to slight changes in their environment, the anticipated changes in global climate are expected to occur too rapidly to be compensated for by normal evolutionary processes such as natural selection. Plant species cannot simply disperse and move north or south into the newly created climatic regions that will be suitable for them. Many tree species take hundreds, even thousands, of years for seed dispersal. Paleobotanist Margaret Davis predicted that in the event of a CO2 doubling, the sugar maple (Acer saccharum), which is presently distributed throughout the Midwestern and northeastern U.S. and southeastern Canada, would die back in all areas except in northern Maine, northern New Brunswick, and southern Quebec. Of course, this contraction in the tree's distribution could be offset by the creation of new favorable habitats in central Quebec. However, most scientists believe that the climatic zones would shift toward the poles faster than trees could migrate via seed dispersal; therefore, extinctions would occur.

We need to be careful in concluding too much from this, though. Species are not the same as individuals, so it could well be that plant species would decline while the number of individuals of the remaining species would increase. For example, maybe the trees wouldn't migrate, but smaller bushy plants would, and these might even grow faster (i.e., convert more sunlight to food) than the trees would have. I don't know -- this question needs further examination; I'm just pointing out that the directional implications for wild populations aren't obvious.

It's also worth noting that even if disruption did decrease populations, it might also cause increased suffering per survivor. Animals have evolved so that they reduce the stress and injury that they endure in their environmental settings. If the environment changes significantly, the animals will on average be more hungry, more incapacitated, etc. and generally will experience more suffering.

### Reduced diversity

Species extinction from climate change doesn't directly speak to how climate change will affect aggregate animal populations, since other species may fill the niches left by species that go extinct. However, reduced species diversity makes ecosystems more fragile and hence more vulnerable to collapse (see "Effects of Climate Change on the Oceans").

### Effects in the ocean

In aggregate, ocean life might matter less than insects on land because zooplankton are generally less sophisticated than terrestrial invertebrates. On the other hand, some estimates put zooplankton abundance higher than insect abundance.

How would climate change affect populations of ocean life?

#### Reduced primary productivity?

One study: "Changes in marine net primary productivity (PP) and export of particulate organic carbon (EP) are projected over the 21st century with four global coupled carbon cycle-climate models. [...] All four models show a decrease in global mean PP and EP between 2 and 20% by 2100 relative to preindustrial conditions, for the SRES A2 emission scenario."

Under future global warming, increasing temperature may stratify the world ocean, decreasing the supply of nutrients from the deep ocean to its productive euphotic zone. Consequently, phytoplankton activity will decline [...].

[...] researchers simulating future oceanic primary production have found evidence of declining production with increasing ocean stratification,[8][9]

#### Possible declines in zooplankton?

Zooplankton are relevant both insofar as they themselves might be capable of suffering and because populations of some larger animals are made possible by them.

##### Studies suggesting zooplankton decline

From Anthony J. Richardson, In hot water: zooplankton and climate change:

Dynamics of plankton communities at a first approximation are captured by nutrient-phytoplankton-zooplankton (NPZ) models. [...] NPZ models can be coupled to [general circulation models] GCMs of the Earth's climate system, allowing investigation of the potential future states of plankton communities under alternative projections of climate.

Results from the NPZ model of Bopp et al. (2004, 2005) suggest that under doubling of pre-industrial CO2 levels, global primary productivity may decline by 5-10%. This trend is not uniform, but indicates productivity increases of 20-30% in high latitudes and marked declines in the stratified tropical oceans (Figure 10). This and other models generally suggest that warmer, more stratified conditions in the tropics will reduce nutrient concentrations in surface waters, which will lead to smaller phytoplankton cells dominating over larger diatoms, thereby lowering zooplankton biomass. [...]

There is already observational evidence supporting some of these model projections. Decreased nitrate availability was apparent in the 20th century during warm periods in both hemispheres, and a decreasing trend is clearly evident globally since the 1970s (Kamykowski and Zentara, 2005). Ocean colour satellite data based on CZCS (1979-1986) and SeaWiFS (1997-2000) show that global ocean phytoplankton chlorophyll decreased 8% from the early 1980s to the late 1990s (Gregg and Conkright, 2002). Behrenfeld et al. (2006) demonstrate that global, depth-integrated chlorophyll biomass since 1999 has dropped by an average of 0.01 Tg year-1. This decline was driven by El Niño-like climatic conditions that enhanced stratification in the expansive stratified low-latitude oceans and consequently reduced nutrient availability for phytoplankton. As some climate models predict more permanent El Niño conditions in a warmer system state, this study suggests that the abundance and productivity of plankton communities in the tropical oceans could decline in the future. There is also some evidence that global time-series of zooplankton abundance are declining in the tropical North Atlantic (Piontkovski and Castellani, 2007). Any future reductions in primary and secondary productivity and export production will not only reduce the food available for higher trophic levels in pelagic ecosystems, but will also impact deep ocean communities (Ruhl and Smith, 2004).

Note that the above study focused primarily on biomass, but I don't know if we can directly translate that to numbers of individuals because warmer temperatures cause copepods and their predators to become smaller in size.

The planktonic copepod Calanus finmarchicus is an important component of the North Sea food web, channelling energy from primary production to harvestable fish resources, and is therefore an indicator of the state of the marine food web. During the 1960s the biomass of Calanus in summer constituted up to 70% of all zooplankton in the northern North Sea, but since then its abundance has declined, so that in the late 1990s its biomass is only around 50% of that found 30 years earlier.

There is a strong statistical relationship between the abundance of Calanus finmarchicus in the northern North Sea and an atmospheric index of climate in the North Atlantic -- the North Atlantic Oscillation (NAO) index. Research at Fisheries Research Services (FRS) has identified the likely oceanographic basis for this relationship as a combination of changing wind patterns in spring, and a steady decline in the volume of cold bottom water in the Faroe-Shetland Channel. In the winter, at depths greater than 600 m, the bottom water contains large numbers (up to 650 m-3) of hibernating Calanus finmarchicus. In the spring these copepods ascend to the surface waters again, and many are carried into the North Sea, maintaining the productive summer stock. The FRS studies indicate that the changes in wind patterns and declining volume of bottom water have effectively reduced the supply of copepods to the North Sea.

From Joseph Kane and Jerome Prezioso, Distribution and multi-annual abundance trends of the copepod Temora longicornis in the US Northeast Shelf Ecosystem:

Temora longicornis abundance in the ecosystem's southernmost subarea (Middle Atlantic Bight) did not increase in the 1990s and was found to be negatively correlated to surface temperature, indicating that continued global warming could adversely impact the copepod[']s annual abundance cycle in this region. [...]

The effects of continued global warming would most likely have the greatest impact on the seasonal abundance cycle of T. longicornis in the southernmost [Middle Atlantic Bight] MAB region. Mean abundance there declines sharply as temperature rises in summer (Fig. 13) and annual levels are lower during warmer years. This negative relationship with temperature is likely caused by a combination of two factors: (i) reduced food concentrations after the spring bloom (Maps et al., 2005; Hansen et al., 2006) and (ii) the deleterious effect of rising temperatures on egg production rates (Peterson and Kimmerer, 1994; Halsband and Hirche, 2001). Long-term data indicates that the range of sea surface temperature on the US northeast shelf is increasing, producing faster warming and cooling rates during seasonal transitions (Friedland and Hare, 2007). Continued warming will likely impact T. longicornis by altering the timing of the spring bloom and raising daily metabolic requirements, lowering rations available for growth and reproduction. Since studies have shown that shelf waters in the region have warmed during the 1990s, with significantly warmer values found during winter (Mountain, 2004; Sullivan et al., 2005), it may not be long before the peak abundance period of T. longicornis is shortened and minimized in the MAB.

That said, this study looked at one species, and I'm not sure how things play out for other zooplankton species in the same habitats.

One slide in a presentation by Carmen García-Comas, Climate change and copepod size spectra: Comparison of two coastal long-term series in the western Mediterranean Sea, notes this trend: Higher water temperature: Lower primary production.

##### Studies suggesting zooplankton increase

Dr Mayor said: "Both of our experiments indicated that the health of copepod eggs remains unaffected when they are exposed to ocean acidification levels predicted for the end of the 21st century. This is great news.

"However our previous research has demonstrated that more severe acidification, potentially arising if a subsea carbon capture reservoir burst open, causes a major decline in the number of copepod eggs that successfully hatch.

"Our most recent study found that the effect of global warming depends on when the eggs were collected.

"In our first experiment we found no clear effect of temperature on how many hatchlings were produced by the eggs.

"But in the second experiment, conducted a week later, increasing the seawater temperature actually increased the number of healthy hatchlings."

Researchers believe this effect relates to the temperature at which the maternal copepods were acclimated -- animals from warmer waters produce eggs that are less stressed by warm water and vice versa.

Dr Mayor added: Our results highlight a potentially positive effect of global warming -- it may increase the number of healthy copepods in our seas, which is good news [sic] for the larvae of fish such as cod and herring, and ultimately fishermen. [sigh :(]

Note that if copepods have more healthy offspring, they can have fewer total offspring to maintain constant populations, so this reduction in r-selection could be a good side effect of what would otherwise be a troubling increase in copepod (and hence, fish) populations.

#### Coral-reef loss

Climate change is possibly the dominant cause of coral-reef degradation, including bleaching and inability to make calcium carbonate. Given that coral reefs are the most productive ecosystems on the planet, it seems likely that what replaces them contains less wild-animal suffering, though it would be good to verify this.

In addition, it's worth noting that the ocean-acidification impacts in particular result primarily from CO2, whereas most of the greenhouse gases emitted by, say, animal agriculture are not CO2. Still: (a) animal agriculture is wasteful of energy in general, hence causing more CO2 emissions than for the average plant-based diet, and (b) coral bleaching is also due to temperature changes, which result from any greenhouse gas, not just CO2.

Ray Hilborn suggests that ocean acidification may (unfortunately) not reduce total productivity:

I don't think either of [ocean acidification or global warming] is a threat to world food production from the oceans because the photosynthetic energy is going to go somewhere, and we'll figure out a way to eat it. But it's going to change things dramatically.

#### Reduced dissolved oxygen

For aquatic animals, one limiting factor is dissolved oxygen. Warmer temperatures due to climate change will reduce dissolved oxygen availability, potentially reducing populations of fish and other animals including copepods.

#### Reduced nutrient cycling

Climate change is expected to make it harder for ocean water to circulate, reducing nutrient cycling and thereby constricting primary productivity.

This page says: "Warming the surface ocean cuts down on mixing with deeper, nutrient-rich water, so phytoplankton productivity drops."

## Effects with unclear sign

### Effects on land

#### Mixed effects on plant productivity

I moved this section to this piece.

#### Changes to water cycle

Climate change is increasing atmospheric water vapor, which might affect plant growth?

Climate change also increases intense flooding, which probably reduces plant growth on balance because the water supply for plants is less slow and steady, and floods can erode soil?

### Effects in the ocean

#### Mixed effects on phytoplankton productivity

This study reports:

We show that under eutrophic conditions, productivity may double as a result of doubling of the atmospheric CO2 concentration. Although in practice productivity increase will usually be less, we still predict a productivity increase of up to 40% in marine species with a low affinity for bicarbonate. In eutrophic freshwater systems doubling of atmospheric CO2 may result in an increase of the productivity of more than 50%. [...] nuisance phytoplankton blooms may be aggravated at elevated atmospheric CO2 concentrations.

While higher primary production usually implies more animal life, this may not be the case in eutrophic waters, so even if the above study is correct, the net impact on overall animal abundances is unclear. That said, if many water bodies are not eutrophic, the net result would be an unfortunate increase in zooplankton and fish populations.

There are several growth-limiting factors for aquatic plants, including light, phosphorus, nitrogen, iron, etc. Global warming may increase ocean stratification in mid-latitudes, decreasing nutrient availability. There's some evidence that this is decreasing chlorophyll density, although other studies find an opposite trend, as discussed below.

In the North and South Pacific, North and South Atlantic, outside the equatorial zone, [...] It is estimated that the low surface chlorophyll areas in these oceans combined have expanded by 6.6 million km2 or by about 15.0% from 1998 through 2006.

[In our model simulations] Climate warming leads to a contraction of the highly productive marginal sea ice biome by 42% in the Northern Hemisphere and 17% in the Southern Hemisphere, and leads to an expansion of the low productivity permanently stratified subtropical gyre biome by 4.0% in the Northern Hemisphere and 9.4% in the Southern Hemisphere. [...] Vertical stratification increases, which would be expected to decrease nutrient supply everywhere, but increase the growing season length in high latitudes.

However, that same study continues:

Four features stand out in the response to global warming: (1) a drop in chlorophyll in the North Pacific due primarily to retreat of the marginal sea ice biome, (2) a tendency toward an increase in chlorophyll in the North Atlantic due to a complex combination of factors, (3) an increase in chlorophyll in the Southern Ocean due primarily to the retreat of and changes at the northern boundary of the marginal sea ice zone, and (4) a tendency toward a decrease in chlorophyll adjacent to the Antarctic continent due primarily to freshening within the marginal sea ice zone. We use three different primary production algorithms to estimate the response of primary production to climate warming based on our estimated chlorophyll concentrations. The three algorithms give a global increase in primary production of 0.7% at the low end to 8.1% at the high end, with very large regional differences.

Oceana's "Effects of Climate Change on the Oceans" suggests that ocean nutrient circulation and hence phytoplankton productivity may decrease with climate change because fresh water from melting ice caps is less dense than sea water.

This section of Wikipedia's "Phytoplankton" article reviews contradictory evidence about whether phytoplankton is increasing or decreasing on balance in the wake of climate change.

#### Mixed effects on plant nutritiousness

Even though algae may grow faster with higher CO2 concentrations, they may become less nutritious.

## Short term vs. long term

Even if climate change reduces animal populations in the short term, it's not obvious this would be the case in the long term as ecosystems have time to adapt to the new conditions. This suggests that if we think the net impact of climate change is currently equally likely to cause fewer as more wild animals in the short term, then in the long term, our expectations would be asymmetric toward more wild animals.

On the other hand, there's some chance ecosystems in general will be less prevalent hundreds/thousands of years hence, in which case the prospect of higher long-term populations would matter less.

## Greater r-selection?

In general, greater environmental instability tends to favor so-called "r-strategists," organisms that have large numbers of children and die at a young age. This piece says: "The vagaries of climate change are such that despite our best efforts, what a particular parcel of natural landscape will look like but a few decades from now is anyone's guess. Instability and uncertainty are the new driving forces when it comes to land management. Such forces create habitats most suitable to r-selection strategies."

This paper says:

Currently, there is little long-term data to indicate how r- and K- selected mammals will respond to climate change. However, a recent study found that frequent severe weather events may drive selection towards precocial maturation in male pronghorn antelope due to high male mortality (Mitchell & Maher 2006), suggesting that climatic variation is driving this population towards a faster life history strategy.

I suspect it's generally bad if K-selected herbivores (e.g., elephants) are replaced by r-selected animals. K-selected animals on average suffer less per unit time than r-selected animals, and there are fewer of them per unit area, so there's less aggregate suffering.

On the other hand, it's less clear whether it's bad to have fewer K-selected omnivores and carnivores, such as large birds. These animals cause painful deaths to large numbers of smaller animals, although a full analysis of their net impact is complex.

In addition, here are two possible arguments why increased environmental instability might be good:

1. If the instability also affects plants, then greater instability plausibly means lower total net primary productivity. With lower plant diversity, niches may remain unfilled, and total plant growth may be lower than what it could be given a more mature successional stage (although the relationship between seral stage and productivity is complex). Lower net primary productivity means less energy with which to create suffering organisms at higher trophic levels.
2. The ultimate r-selected life forms are bacteria and other unicellular organisms, which we may care about less per unit of metabolism than we care about small animals. Instability that increases numbers of these kinds of organisms, and crowds out animal decomposers, may be good.

### Reduced body size

Warmer temperatures from climate change may reduce animal body sizes. This might lead to

1. greater populations, because the same amount of energy in an ecosystem would be divided over smaller animals
2. increased r-selection because the intrinsic rate of increase of a population seems empirically to increase when body size decreases and when temperature increases.

However, organisms may get smaller to compensate for a faster metabolism. For example, springtails don't grow bigger and may shrink at high temperatures because higher temperatures imply higher metabolisms (energy used per unit time per gram of body mass), which needs to be offset by reducing the animal's total body mass. If the shrinkage exactly compensates for the increased metabolism per gram (does it?), then the total energy used by these smaller organisms would be the same as the total energy used by the same number of bigger organisms with slower metabolisms. Hence, shrinking body sizes doesn't necessarily mean more total animals can be supported by a given amount of energy.

This paper reports:

An increasing body of evidence suggests that climate change has already resulted in significant changes in body mass and size in a number of mammal species. According to Bergmann’s (1847) Rule, animals in warmer areas should be smaller than those in cooler areas, as the lower surface area:volume ratio of larger animals may assist in heat conservation in cooler climates. Thus, it could be expected that climatic warming will result in a decrease in body size and/or mass, and a number of studies have demonstrated that this has occurred in some mammals and birds (e.g. Smith et al. 1998, Yom-Tov 2001). However, the converse pattern has also been found in masked shrews Sorex cinereus in Alaska and otters Lutra lutra in Norway (Yom-Tov & Yom-Tov 2005, Yom-Tov et al. 2006a), and in some other vertebrates (Yom-Tov 2001, Yom-Tov & Yom-Tov 2005, Chamaille-Jammes et al. 2006).

Studies of 17 endothermic vertebrates (including 4 mammals) in which a clear change in size or mass was demonstrated (Smith et al. 1998, Yom-Tov 2001, YomTov & Yom-Tov 2004, 2005, Fernandez-Salvador et al. 2005, Yom-Tov et al. 2006a,b) showed that 71% (n = 12) of the vertebrates investigated declined in body mass. For those studies where latitude was reported, the latitude of species which declined in mass ranged from 31 to 52° N, while the latitudinal range of species which showed an increase in mass was 37 to 65° N. Thus, while the prevailing pattern of a decline in body mass conforms to the predictions of Bergmann’s Rule, it is notable that some species, particularly those at higher latitudes, show the reverse pattern. At more northerly latitudes and/or at high altitudes, global warming may have decreased winter severity and resulted in an increase in food availability, allowing species a longer growth period and the attainment of higher body mass.

## Delaying a future ice age

Some scientists believe that the Earth would return to an ice age in roughly 1500 years if humans were not elevating CO2. Anthropogenic carbon emissions may delay the time until the ice age returns, unless humans change the situation in the meanwhile, such as by removing CO2 from the atmosphere.

I suspect that if human technology continues at its current pace, most intelligence will be digital in 1500 years, and wildlife as we know it may not exist outside of preserves and laboratories. In this case, climate change wouldn't have much of an impact on wild animals that far down the road. But there's a reasonable chance that civilization will be set back technologically and a tiny chance that humans will go extinct altogether, and in these cases, the effects of our carbon emissions on life 1500 years hence would still matter. And from the perspective of wild-animal suffering, those effects would be clearly negative, since forestalling an ice age would keep flora and fauna populations far higher than they would be during an ice age, thereby causing vastly more suffering. I don't know the exact relationship between CO2 emissions and number of years of delay in resuming an ice age, but I expect the effect is nontrivial. In the BBC article, Lawrence Mysak predicted: "Absorption by the oceans takes thousands or tens of thousands of years - so I don't think it's realistic to think that we'll see the next glaciation on the [natural] timescale."

Peter Sinclair expresses skepticism regarding the exact prediction of 1500 years: "The parameters of these orbital changes have been fuzzy enough that estimates for the onset of the next ice age are kind of all over the map, and I’ve seen estimates from 1000, to 30,000 years in the future." But he agrees that the impacts of CO2 are likely to be long-lasting:

The study’s bottom line is that, for a glaciation to occur, carbon dioxide levels would have to get back down to 240 parts per million (ppm), from the current 390. Given the very slow drawdown of co2 from natural processes, even if we were to stop emitting today, that would not be happening for some millennia in the future.

## Anthropocentric considerations argue against climate change

Climate change is expected to cause extensive harm to poor humans, through greater crop-yield variability, wider spread of disease, rising sea levels, and other factors. Many people care a lot about preventing climate change, and as a general heuristic, if you can do something that other people want at low cost to your values, you should do it, in return for goodwill and reciprocity benefits. In light of the major uncertainty about the impact of climate change on wild animals, this suggests we should aim to reduce our greenhouse-gas emissions where it's easy for us to do so.

## Don't buy biomass-based carbon offsets

Even if we think climate change is net bad in expectation, I don't support buying many types of carbon offsets. For one thing, there's uncertainty about whether the promises made by sellers of carbon offsets will be kept, though third-party verified offsets may be more trustworthy. But even if the carbon offsets are genuine, they often involve preserving or planting forests in order to sequester carbon, which is bad for wild animals.

In particular, consider the case of Cool Earth, which preserves rainforest land. According to Giving What We Can's report on the charity, preserving one acre of rainforest can save 260 metric tons of CO2, or 640 metric tons per hectare. Giving What We Can says this is probably an underestimate, because it doesn't count soil carbon stores, nor does it count non-CO2 greenhouse gases that are released by deforestation. So I'll round the figure up to (just making this up) 800 metric tons per hectare.

Let's modify the calculation at the beginning of this piece with some more realistic numbers. Suppose there's a 55% chance that climate change will increase complexity-weighted insect/zooplankton populations by 5% on balance and a 45% chance it'll reduce them by 5% on balance. The net expected effect is to increase by 5% * (55%-45%) = .5%. Then, saving one hectare of rainforest prevents 800 metric tons of CO2 and thereby avoids this many wild insect-years:
$\frac{800\:\textrm{tons}}{34.5\cdot10^9\:\text{tons}}\:\cdot.5\:\text{percent}\cdot\:10^{18}\:\text{insects}\:\cdot\:1\:\text{assumed}\:\text{year}\:\text{of}\:\text{impact}=\\\\100\:\text{million}\:\text{insect-years}$

In contrast, how many insect-years will be created by preserving a hectare of rainforest indefinitely? Crop land may be home to tens of millions of insects per hectare, so rainforest probably has even more. If rainforest has, say, 20 million more insects per hectare on average than deforested land (which seems like a low estimate), then 5 years of preserving the hectare of rainforest would be as bad as the expected benefits from averting climate change are good. Given that the average insect density over the planet is 75 to 750 million insects per hectare, with rainforests probably ranking near or at the top of the list for insect density, it's quite possible that there are at least hundreds of millions of insects in a hectare of rainforest land, in which case saving a hectare of rainforest land may be net bad even within one year, assuming deforested land supports fewer insects.

So far I've been assuming that the rainforest land would be preserved indefinitely, though this probably wouldn't be the case, since it's hard to maintain promises not to use land indefinitely in poor countries. If the land is eventually harvested and the carbon released, then the CO2 sequestration benefits would be lost at that point. This would exacerbate climate change later on, though carbon is somewhat less bad further in the future either (1) because humans may have better ways of capturing it or otherwise reversing climate change by then or (2) because the Earth will have less biological life in general by then, so that the raw increase in insect populations from climate change will be smaller. On the other hand, the sooner the rainforest land is cleared, the sooner the insects that would have lived on it avoid the harm of being born.

Not all carbon offsets work by preserving forests, and those that don't might be okay (if not necessarily cost-effective) from a wild-animal perspective. For example, some of Google's carbon offsets involve burning methane. But buying non-forest-based offsets may shrink the pool of non-forest carbon-offset opportunities, leading other people to buy more forest-based carbon offsets, which would be bad. Or maybe buying more non-forest offsets increases the market size and economies of scale in non-forest offset opportunities, which would be good. Also, supporting companies that do offsets would probably in the long run increase the number of biomass-based offsets used, which would be bad.

This argument against carbon offsets doesn't apply to efforts to reduce your personal carbon emissions (unless such efforts involve preserving rainforest). It's probably net good in expectation to save energy in your home, drive/fly less, and so on.

Rather than buying carbon offsets, it might save at least as much carbon per dollar to donate to a charity that advocates for energy efficiency and renewable energy, since advocacy in general has higher expected returns than direct charity. One example 501(c)(3) charity that looks promising is the Center For Energy Efficiency And Renewable Technologies (CEERT), since it basically only does efficiency improvements and renewables, not habitat/biomass preservation. Rather than buying $100 of carbon offsets, consider donating$100 to CEERT. Or look for a charity that lobbies for funding of research on nuclear fusion or other high-risk, high-return gambles. Don't donate to a generic environmentalist charity, since they undertake efforts to preserve wilderness in addition to pushing energy efficiency and renewables.

## Acknowledgements

Thanks to several people, including Max Maxwell Brian Carpendale and Jonathan Erhardt, for pointing me to relevant information for this article.