Connectivity at a cost: Economic dynamics of restoring habitat connectivity

Natural Resource Modeling. 2020;e12294. wileyonlinelibrary.com/journal/nrm

|

Connectivity at a cost: Economic dynamics of

restoring habitat connectivity

Wisdom Akpalu

|

Jesper Stage

1

School of Research and Graduate Studies, Ghana Institute of Management and Public Administration, Achimota‐Accra, Ghana 2

World Institute for Development Economics Research, United Nations University, University of Ghana, Accra, Ghana

3

Department of Business Administration, Technology and Social Sciences, Luleå University of Technology, Luleå, Sweden 4

Mid Sweden University, Sundsvall, Sweden

Correspondence

Jesper Stage, Department of Business Administration, Technology and Social Sciences, Luleå University of Technology, 971 87 Luleå, Sweden.

Email:Jesper.Stage@ltu.se

Funding information

Elforsk; The Swedish R&D programme Kraft & Liv i Vatten

Abstract

Both in the United States and in Europe there is ongoing work on reversing habitat fragmentation and the attendant loss in biodiversity in river systems caused by hydropower and other developments. Fish ladders and other measures are being introduced to restore the connectivity in river systems. In this pa-per, we set up a theoretical model to investigate what the conditions are for such an investment to be so-cially profitable. We find that, even in cases where it would have been socially preferable not to build hy-dropower installations in the first place, connectivity‐ restoring measures affecting the installations are not necessarily socially beneficial. This is the case for a wide range of plausible assumptions about discount rates, investment costs and productivity losses.

Recommendations for Resource Managers: • Even in situations where it would have been

so-cially preferable not to build hydropower plants in a river, carrying out measures now to reconnect different parts of the river and re‐establish ecolo-gical connectivity may not be socially desirable.

- - - --This is an open access article under the terms of the Creative Commons Attribution‐NonCommercial‐NoDerivs License, which permits use and distribution in any medium, provided the original work is properly cited, the use is non‐commercial and no modifications or adaptations are made.

© 2020 The Authors. Natural Resource Modeling published by Wiley Periodicals LLC Wisdom Akpalu and Jesper Stage contributed equally to this study.

• Any economic analysis of the value of re‐ establishing ecological connectivity needs to con-sider the time lags involved until the river ecology is restored.

• Reducing the time lags by actively resettling species does not necessarily improve the overall social profitability of reconnecting the different parts of the river.

K E Y W O R D S

ecological connectivity, habitat fragmentation, hydropower, species‐area relationship

1

|

I N T R O D U C T I O N

Hydropower development in the Western world during (and before) the 20th century con-tributed significantly to economic development by providing a reliable and relatively cheap source of electricity in many countries (see Gan, Eskeland, & Kolshus,2007). However, the environmental effects including biodiversity loss, which are important, were often not fully accounted for (Isbell et al.,2013; Kaunda, Kimambo, & Nielsen,2012). Since the hydropower plants were typically built with little consideration for environmental and ecological impacts, the new plants destroyed numerous freshwater habitats and cut off important migratory routes for fish and other freshwater species (Chen et al., 2016; Cheng, Li, Castello, Murphy, & Xie,2015; Cooper, Infante, Wehrly, Wang, & Brenden,2016; Mueller, Pander, & Geist, 2011; WWF,2004). In recent decades, therefore, there have been efforts to restore habitats adjacent to hydropower plants and to restore ecological connectivity between different parts of river sys-tems that have been separated by plants and dams (Hart et al., 2002; Stage,2018). However, although some ecological and biological studies have been made of these restoration efforts, the economic literature on this topic has been quite limited. This is unfortunate, as any restoration measures will tend to be long‐term in nature and, thus, the costs and benefits of carrying out such measures deserve more attention from economists.

In this paper, we explore the economic implications of restoring ecological connectivity in a river system that has been subdivided ecologically into smaller fragments by hydropower in-stallations. We find that, even under assumptions that are likely to overstate the economic benefits of restoration measures, there is a wide range of economically and ecologically plau-sible parameter values for which restoration measures are not economically beneficial. Notably, the results remain consistent for a range of plausible parameter values, including a scenario where the economic benefits from a functioning river ecosystem are such that the river should not have been developed for hydropower in the first place.

We begin the paper by discussing the ecology of habitat fragmentation—not only in river habitats, but also more generally. We proceed by setting up a simple baseline model to illustrate some of the key economic issues involved. We then introduce a more realistic model where additional real‐world complications associated with habitat connectivity restoration are brought

into the modeling, and show that many of the key results remain—and, if anything, are strengthened further. In a subsequent section, we bring in an additional complication: the possibility of speeding up the ecosystem's recuperation through resettlement measures. We find that this does not affect any of the key results in the model. We conclude by discussing implications of these results for policy and for further research.

2

|

H A B I T A T F R A G M E N T A T I O N

Many natural habitats have become increasingly fragmented with time, due to various human activities (Andren, 1994; Powers & Jetz, 2019; Wilcove, McLellan, & Dobson, 1986). Such

fragmentation has frequently led to reduced biodiversity (see e.g., Andren, 1994;

Diamond, 1972; Diamond & May, 1981; Harris, 1984; MacArthur & Wilson, 1967; Wearn, Reuman, & Ewers,2012), with attendant losses in both use and non‐use values, for example,

reduced existence values caused by extinction or increased scarcity of valued species (Mitchell et al.,2015). While the economic activities causing the fragmentation have frequently generated important economic benefits for some agents in society, the welfare losses linked to the con-comitant habitat fragmentation and loss of biodiversity have been important too, both for people in the surrounding areas and for the global community (Hanski, 2011). Accordingly, policies aimed at reconnecting the remaining habitats have also been implemented (Xu et al.,2006). For land habitats, however, the problem has frequently continued to worsen over time: in many cases, increased growth in the economic activities competing for land has led to ongoing encroachment on the affected habitat areas.

For river ecosystems in the developed world, the situation is slightly different from that on land. For many of the fragmented river ecosystems and river habitats in Europe, the United States, and elsewhere, the fragmentation took place during the 1900s and is no longer an ongoing process. In particular, hydropower developments cut off the migration routes for nu-merous fish species, leading to population loss and even the local extinction of many of them. Such developments also had important negative impacts on species without clear migration patterns: they lost access to nutrient flows and lost parts of their natural habitats (Ziv, Baran, Nam, Rodríguez‐Iturbe, & Levin, 2012). Therefore, both national policies and supranational policies such as the European Union Water Framework Directive have been put in place in recent decades to improve the health of aquatic ecosystems (see, e.g., O'Connor, Duda, & Grant, 2015). While there may be special interest in one or a few species (see, e.g., Håkansson,2009, for salmon, Paulrud & Laitila,2013, for sea trout, or Stage,2015, for eel) the policy goal is usually to restore biodiversity per se, not only one or a few specific species. Thus, guidelines for implementing the European Water Framework Directive (see e.g., European Commission, 2015) stress improved connectivity and aquatic biodiversity per se rather than outcomes for individual species. This is because it is widely recognized that a healthier and more diverse ecosystem can also more easily support the keystone species (see e.g., Akpalu & Bitew,2011). When effects on a specific species are monitored, this is often because that species is seen as a useful indicator of overall ecosystem health rather than because other species are seen as unimportant. Thus, unlike the situation on land, where the primary goal is to stop the habitat fragmentation from getting worse and to prevent ongoing losses of biodiversity, the policy goal for many of these river systems is one of trying to shift from the current steady state to one that would be better for the river ecosystems, through restoring connectivity that has been lost for decades and recovering biodiversity that has already been lost.

Ecologists studying habitat fragmentation have established a species–area relationship, which gives the equilibrium number of species Seqsupported by a given habitat size A,

S =cAz.

eq (1)

In this relationship, c is a constant and z is a constant between 0 and 1. The exact value of z varies by type of habitat, but tends to be more or less constant between different habitats of the same type (see e.g., Brooks & Balmford,1996; Drakare, Lennon, & Hillebrand, 2006; Pimm, Russel, Gittleman, & Brooks,1995; Rosenzweig, 1995; for river and freshwater systems speci-fically, see e.g., Angermeier & Schlosser,1989; Eadie, Hurly, Montgomerie, & Teather,1986).

This species–area relationship means that, if a habitat is reduced in size or if it is subdivided into several unconnected habitats, the new, smaller habitat(s) will see the number of species S decline over time until a new and lower equilibrium level is reached. Conversely, it also implies that, in the less frequent case where fragmented habitats are reconnected, the new habitat will be able to support a higher equilibrium number of species; the number of species in the habitat will then increase over time as the diffusion of animals and plants from other habitats leads to new (and old) species (re‐)establishing themselves in the restored habitat.

Wearn et al. (2012, in line with research by, for example, Diamond et al.,1972) report on a study on local extinction rates in fragmented habitats. The authors find empirical evidence suggesting that the rate of change in the number of species, when the available habitat size has changed, can be described by the simple differential equation

dS

dt = (k cA − ),S

z ₍₂₎

where the new habitat size A determines the new equilibrium number of species, S is the actual current number of species, and k is a parameter measuring the speed at which the number of species adjusts to the new equilibrium level. The value of k appears to be quite low in many cases; Wearn et al. (2012) estimate a mean value for k at 0.0122, implying that it takes 57 years for half of the loss in species number to occur, and the earlier study by Diamond (1972) suggests kvalues of 0.0004 or less. Similarly, recovery in species numbers when habitats are reconnected also appears to take a relatively long time: efforts to reconnect previously fragmented river habitats by installing new fish bypasses and fish ladders began in earnest in the 1980s, but species diversity has yet to recover fully in many of the river systems where this was done (see e.g., Hasselquist et al.,2015).

This dynamic relationship between the available habitat and the rate of change of the number of species allows us to compare the economic benefits of restoring habitats—linked to improved economic benefits from the increased number of species—to the cost of restoration efforts. The following section sets up a simple baseline model for this comparison.

3

|

A S I M P L E B A S E L I N E M O D E L

Assume that a river habitat A (the river itself and the adjacent beach zone) has been subdivided into N fragments by N− 1 hydropower plants. To keep the model as simple as possible, we assume that all N habitat fragments are of equal size. Furthermore, let the hydropower plants be identical and produce equal continuous quantities of hydropower, worth h each. Because of the

hydropower development, the equilibrium number of species in each habitat fragment has declined fromS =cAz eq to S c A N cN A = = . z z z eqnew ⎜⎛_⎝ ⎟⎞_⎠ − (3)

Assume that hydropower development took place a long enough time ago for the new equilibrium species number to have been reached. Assume further that the N habitat fragments all have the same remaining species, such that biodiversity loss has been maximal, and overall species loss is cAz(1 −N−z)_{, that is (S −S}

eq eqnew). Although this latter assumption is likely to

exaggerate the overall loss, as the overlap in species between different habitat fragments is unlikely to be complete, it simplifies the subsequent modeling. Note that since each habitat will have the same species, the term on the right‐hand side of Equation (3) is not multiplied by the number of habitats.

Also for simplicity, we assume that the existence value of biodiversity (B) in the river system is a linear function of S, B S( ) =bS. (A more realistic valuation function might be concave rather than linear in the number of species, since linearizing exaggerates the value of additional species.)

Assume that it is possible to restore perfect connectivity between the N patches of habitat by building bypasses that cost I each in initial investment cost and entail a continuous cost of a (which is a fraction of the overall hydropower value, 0 < a < h, and which may be due to reduced production, shifts to producing at less profitable times, or both) each, giving the following present value of the total cost function:

(

)

∫

From an economic perspective, such a restoration programme is justifiable if species numbers recover quickly enough (through colonization from habitats elsewhere). Thus,

bS t e dt N I N h a e dt bcN A e dt N he dt ( ) − ( − 1) + ( − 1) ( − ) > + ( − 1) . rt rt z z rt rt 0 − 0 − 0 − − 0 −

∫

∫

∫

∫

The left‐hand side of the inequality gives the overall value generated by the river system if connectivity is restored, with the first term showing the present value of the recovering bio-diversity, the second term showing the investment cost, and the third term showing the present value of the remaining hydropower generation. The right‐hand side gives the overall value generated by the river system in its present state, with the first term showing the present value of the remaining biodiversity and the second term showing the present value of the hydropower generation in the system's current state. The recovery rate of species diversity, if connectivity is restored, is given by Equation (2)

(

)

dt

z _with

S(0) = cN A−z z, ₍₆₎

S t( ) =cAz(1 + (N−z − 1)e−kt). ₍₇₎

Restoring connectivity is obviously not economically beneficial if the costs outweigh the benefits. This happens if we have (using 7 in 5)

bcA N k r k N rI a (1 − ) + < ( − 1)( + ). z −z _⎜⎛ _⎟ ⎝ ⎞ ⎠ (8)

Thus, the connectivity restoration may be too costly if the social opportunity cost (given by the annualized investment cost plus recurrent expenditure, multiplied by the number of power plants, that is,(rI+ )(a N− 1)), is sufficiently high to offset the social gains from improved

biodiversity – which comprises of the net equilibrium gain from restoration (i.e.,

bcAz(1 −N−z)_{), discounted at an effective discount factor}

( )

r+k . There is no a priori constraint on I or a, so this may well be the case in river systems where hydropower production is valuable or where biodiversity would be sparse in any case. On the other hand, an elevated speed of adjustment, denoted by a higher value ofk, could potentially make (otherwise unprofitable) connectivity restoration profitable.

That the cost of restoring connectivity might be greater than its benefits is, to some extent, trivial: since it was considered worthwhile to develop the river system in the first place, the net benefits of doing so must clearly have been believed to be positive. To provide for more in-teresting results, let us assume that the benefit of the hydropower is less than the existence value of the lost biodiversity evaluated at the equilibrium number of species, meaning that it would have been better if the hydropower development had never occurred. Thus,

h N( − 1) <bcAz(1 −N−z). ₍₉₎

This could be the case if the biodiversity in the river system is very valuable, if the hy-dropower in the river system is not particularly valuable, or both. That development has taken place despite this might be, for example, due to undervaluation of biodiversity, which may occur if the ecological scarcity of the species is not fully captured (Barbier,2012). This can make what seemed like a justifiable trade‐off when the river was developed no longer defensible today. Furthermore, the development may not have made economic sense in the first place, if legislation did not take existence values into consideration when the river was developed, thereby imposing an externality on the rest of society. If (9) holds, it is also the case that,

N rI a N rI N a N rI bcA N a

h

( − 1)( + ) = ( − 1) + ( − 1) < ( − 1) + z(1 − −z) . ₍₁₀₎

For connectivity restoration to be less beneficial than maintaining the river system in its present state, we would need the leftmost inequality in

bcA N k r k N rI a N rI bcA N a h (1 − ) + < ( − 1)( + ) < ( − 1) + (1 − ) z −z _⎜⎛ _⎟ z −z ⎝ ⎞ ⎠ (11)

to hold; the rightmost inequality holds by assumption from (10). The two inequalities can hold simultaneously if the investment cost I associated with building a bypass is sufficiently high, and/or if the effective discount factor of the net benefit from the biodiversity recovery is very

low due to low speed of adjustment of species to the long run equilibrium levels or a high social discount rate. A low speed of adjustment or relaxation rate signifies a high biodiversity recovery time lag. Thus, if the time lag is too long, the present value of the lost hydropower (which begins to generate costs for society as soon as the connectivity restoration programme is im-plemented) may outweigh the present value of the recovering biodiversity (the full value of which will only become available to society after a time lag).

In some cases, typically for large‐scale plants where h is high, the investment costs of establishing bypasses will be relatively low in comparison with the overall costs of a con-nectivity restoration programme. If we take the limit of this and set I = 0, concon-nectivity re-storation may still be suboptimal if Equation (12) holds:

bcA N k r k N a bcA N a h (1 − ) + < ( − 1) < (1 − ) . z −z _⎜⎛ _⎟ z −z ⎝ ⎞ ⎠ (12)

This may occur if

k r k a h + < . ⎜ ⎟ ⎛ ⎝ ⎞ ⎠ (13)

Equation (13) implies that, even when the equilibrium biodiversity gain exceeds the loss in hydropower due to the restoration, it is still not economically viable to restore if the effective discount rate is lower than the restoration cost per unit of revenue generated from hydropower. To examine how likely this is in practice, let us consider possible empirical values for the different parameters involved. k appears, empirically, to be quite low. Wearn et al. (2012) derive a mean k value of approximately 0.0122 from their data, Diamond (1972) derived even lower values, and as noted previously, recovery times in river habitats in the developed world have also been quite long. Typical r values range from 0.02 to 0.05 in developed countries. With Wearn et al.'s (2012) value for k and with r = 0.05, connectivity restoration can entail social losses even if the share of hydropower a/h that is lost is only 0.20, and with lower values of k even smaller losses would be required to make reconnection unprofitable.

Actual empirical estimates of a/h, the share of hydropower lost if connectivity is restored, are harder to come by; as noted in the introduction, there are surprisingly few economic assessments of measures to restore connectivity. However, there are gray literature estimates from several European countries that can be used as an indication. Austrian estimates (Baumann & Lang, 2013; Stigler, Huber, Wulz, & Todem, 2005) suggest that for the country's hydropower sector as a whole, the measures needed to meet the requirements of the Water Framework Directive would only entail losses of, on average, 4% of hydropower production. However, an important reason for these low average losses is that this national average includes pump storage plants with limited impact on connectivity. For the country's small‐scale hydropower plants, the average losses would be 12% of production, and for these plants it should be noted that the assumption that I can be ignored is less realistic than it is for larger plants. In Switzerland, similar estimates (Kummer, 2002), indicate that production losses would be 6% on average for the country as a whole, but would reach 30% in some plants. In Sweden, legislation has limited production losses to at most 5% of production in most renegotiations of hydropower concessions, but this has been seen as insufficient for restoring ecological health in most cases, suggesting that the production losses needed to restore the previous state of the ecosystem are usually higher. Hedenskog and Monsén (2012), in a review of concession renegotiations, identified a number of

cases where environmental courts had decided to require production losses above 5%, and in some cases, production losses of up to 25% were considered reasonable to improve environmental quality in the river system in question. Bergsten et al. (2014), studying one river system with small scale plants and one with large scale plants, found that far greater losses than 5% would be needed in both river systems to restore the previous river ecology—effectively, removing entire plants would be necessary in both systems.

Thus, under relatively plausible assumptions about recovery rates and discount rates, under relatively generous assumptions about the value of restored biodiversity, and with extremely optimistic assumptions about investment costs, in river systems where hydropower develop-ment should never have been carried out in the first place it may nonetheless be suboptimal to restore connectivity if this entails losing more than a fifth of the hydropower currently being generated. Moreover, anecdotal evidence from gray literature in several countries suggests that situations where this type of loss is necessary to restore connectivity do in fact occur regularly.

4

|

B R I N G I N G M O R E R E A L I S M I N T O T H E M O D E L

The model set up in Section3can be made more realistic in many ways. Given experiences with actual hydropower developments, it seems reasonable to assume that, if hydropower remains in use, water flow patterns will be different—with adverse effects on ecosystems in and sur-rounding the river. As a result, apart from the loss of habitat connectivity, the overall avail-ability of habitat will be lower than it was before the river was developed for hydropower. This habitat will continue to be lost if hydropower remains in use, which means that even a river system where connectivity has been restored will not see the entire previous habitat quality restored. This loss of overall habitat availability needs to be considered in a more realistic model. One way of doing so would be to let the sizes of the patches of habitat after hydropower development be given byA μh

N

−

each, rather than byA

N, allowing the habitat loss in each patch to be linked to the amount of hydropower produced. Note that μ is a conversion factor that translates revenue from power generation to area covered by the hydropower plants. With this revised expression, the total biodiversity available is lower than in the simpler model, both with and without connectivity restoration.

On the other hand, given the lengthy delays before the full impacts on species richness is realized, it may not be realistic to assume that the river system has already reached its frag-mented steady‐state species level. Wearn et al. (2012), studying fragmented habitats in the Amazon, find that much of the species loss implied by current fragmentation levels has yet to occur there. In their view (Wearn et al.,2012),“This time delay offers a window of conservation opportunity, during which it is possible to restore habitat or implement alternative measures to safeguard the persistence of species that are otherwise committed to extinction.” The same argument could apply to river habitats fragmented by hydropower, even decades after such fragmentation occurred.

Incorporating these considerations into the model gives us that, in its current state, the developed river system will generate a value function of

bS t N h e dt

( ( ) + ( − 1) ) rt ,

0

−

∫

dS

dt = (k cN (A−μh) − )S

z z

− ₍₁₅₎

and where the species richness S(0) is higher than the equilibrium level to begin with. Solving for S(t) gives us

S t( ) =S(0)e−kt +cN−z(A−μh) (1 −z e−kt).

(16) Integrating the value function in (14), we can therefore write

b S cN A μh r k cN A μh r N h r (0) − ( − ) + + ( − ) + ( − 1) , z z z z − − ⎛ ⎝ ⎜ ⎞_⎠⎟ (17)

if the river system is maintained in its present, developed and unconnected, state.

On the other hand, if connectivity measures are carried out at time zero, the corresponding value function from the river system will be given by

bS t N h a e dt N I

( ( ) + ( − 1)( − )) rt − ( − 1) .

0

−

∫

The equation of motion for species richness will now be given by dS dt = ( (k c A−μh) − ).S z ₍₁₉₎ This gives us S t( ) =S(0)e−kt + (c A−μh) (1 −z e−kt). (20) Using (20) in (18) gives the following value function for the reconnected river system:

b S c A μh r k c A μh r N h a r N I (0) − ( − ) + + ( − ) + ( − 1)( − ) − ( − 1) . z z ⎛ ⎝ ⎜ ⎞_⎠⎟ (21)

Comparing (17) and (21), we see that restoring connectivity is economically harmful if

bc A μh N k r k N rI a ( − ) (1 − ) + < ( − 1)( + ). z −z _⎜⎛ _⎟ ⎝ ⎞ ⎠ (22)

Two of the additional factors that were included in the analysis, namely the continued loss of habitat because of hydropower development and the maintenance cost of the bypasses, both serve to reduce the social profitability of restoring connectivity. The third additional factor, namely the initial non‐steady‐state number of species, appears in the same form in both value functions; this factor therefore cancels out and does not affect the results. Thus, the conclusions from Section3remain: there will be a range of economic parameters (even greater in this more realistic version of the model) for which restoring connectivity is economically suboptimal, even

in situations where the river system should not have been developed for hydropower in the first place.

5

|

A R T I F I C I A L R E S E T T L E M E N T

We saw in Sections 3 and 4 that k, the parameter determining how quickly species

diversity adapts to changes in the equilibrium, is key for determining whether restoring connectivity is socially desirable or not, with lengthy recovery times making restoration less attractive. To speed up diversity recovery, resettling species artificially might there-fore seem an attractive option. Rather than wait for species to colonize the reconnected habitat of their own accord, one could transfer many of them from habitats where they still thrive to the reconnected river habitat, or breed them in captivity for future transfer to the habitat concerned.

However, somewhat surprisingly, artificial resettlement turns out not to have any effect on the results. To see this, let us assume that resettling species entails a fixed cost f per resettled species. In the simpler version of our model, the net benefit (value function) of resettling S′species artificially in a reconnected river system is given by

bS t e( ) rtdt − (N− 1) + (I N− 1) ( − )h a e rtdt −fS′, 0 − 0 −

∫

∫

The associated equation of motion is (2), and the corresponding solution is

S t( ) =cAz(1 + (N−z − 1)e−kt) + ′S e−kt. ₍₂₅₎ Thus, the net benefit, compared with merely restoring connectivity, is given by

bS e′ k r tdt−fS′ ( + )k r bS′ −fS′.

0

−( + ) −1

∫

From Equation (26) it is economically beneficial to reintroduce new species if the discounted marginal ecological benefit exceeds the marginal cost of introducing it (i.e.,

k r b f

( + )−1 > _{). However, the net benefit declines over time because the number of species} would have recovered over time in any case. With a higher number of species to begin with, the rate at which the number of species grows—and, hence, the rate at which benefits from species diversity increases—will diminish.

However, resettling species without restoring connectivity would also give a net benefit, namely of

bS e′ k r tdt−fS′.

0

−( + )

Here, the decline in the net benefit over time is due to the fact that the species will eventually disappear from the river system again. The rate at which the number of species declines is given by dS dt = (k cN A − )S z z − ₍₂₈₎ with S(0) =cN A−z z + ′,S ₍₂₉₎ giving us S t( ) =cN A−z z + ′S e−kt. ₍₃₀₎

Apart from this, however, we see that the net benefit of resettling species is the same— whether connectivity has been restored or not. The more complicated model in Section4yields analogous results.

6

|

C O N C L U D I N G D I S C U S S I O N

Reconnecting river habitats that have become fragmented as a result of hydropower and other anthropogenic interventions is currently an important policy issue in many Western countries. Work continues in numerous river systems, and so does biological research on the issue. Our results suggest that more research is needed on the economic aspects as well.

Not surprisingly, our results indicate that, for some river systems, the costs of restoring connectivity may be prohibitively high in terms of the benefits that can be expected. More unexpectedly, even with a fairly simple model, we find that the costs may outweigh the benefits even for river systems where it is now clear that hydropower should never have been estab-lished in the first place. This result remains even when the model is made more realistic (if anything, this strengthens the conclusion) and when the possibility of artificial resettlement of species is brought into the analysis. Thus, even when it is clear that keeping a pristine, un-disturbed river system would have been a better option, work on restoring habitat connectivity should still be subjected to a thorough cost–benefit analysis which carefully considers the expected recovery rates of the affected species.

There is an important literature in environmental economics dealing with the economics of irreversible decisions (see e.g., Pindyck,1991). The biodiversity loss modeled in our paper is not irreversible in a biological sense, since the species diversity can recover once the habitat is reconnected. Nonetheless, the loss is effectively irreversible in an economic sense. Once the loss has occurred, the economic logic may be in favor of keeping the new, poorer‐diversity state, even in a situation where there is widespread agreement that it would have been better if the loss had not occurred in the first place. The choice facing society in that situation is, effectively, between wasting social resources on restoring diversity that is not valuable enough to justify the expense, or maintaining a status quo that most people agree is worse than the pristine state. Where habitat fragmentation continues to occur, in river habitats in poorer countries where hydropower development is still ongoing and in land habitats throughout the world,

policymakers should do their best to maintain connectivity so as to avoid leaving future generations with that unpleasant choice.

A C K N O W L E D G M E N T S

The authors gratefully acknowledge numerous constructive comments and criticism on earlier versions of this study from two anonymous reviewers; seminar participants at the Centre for Environmental and Resource Economics in Umeå, Sweden, at the Department of Economics, Swedish University of Agricultural Sciences, and at the Department of Economics and Finance, University of Wyoming; and participants at several conferences. Financial support from Elforsk (now part of Energiforsk) as well as from the R&D programme Kraft och liv i vatten (KLIV) supported by several hydropower companies in Sweden, the Swedish Energy Agency, the Swedish Agency for Marine and Water Management and Sweden's water authorities is grate-fully acknowledged. Sandie Fitchat provided valuable help with language editing. The usual disclaimers apply.

O R C I D

Wisdom Akpalu http://orcid.org/0000-0003-2779-2899

Jesper Stage http://orcid.org/0000-0001-7206-6568

R E F E R E N C E S

Akpalu, W., & Bitew, W. (2011). Species diversity, fishing induced change in carrying capacity and sustainable fisheries management. Ecological Economics, 70, 1336–1343.

Andren, H. (1994). Effects of habitat fragmentation on birds and mammals in landscapes with different proportions of suitable habitat: A review. Oikos, 71, 355–366.

Angermeier, P. L., & Schlosser, I. J. (1989). Species‐area relationship for stream fishes. Ecology, 70(5), 1450–1462. Barbier, E. B. (2012). Natural capital, ecological scarcity and rural poverty, Washington, DC: The World Bank. Baumann, M., & Lang, B. (2013). Entwicklung energiewirtschaftlicher Inputdaten und Szenarien für das

Klimatschutzgesetz und zur Erfüllung der österreichischen Berichtspflichten des EU Monitoring Mechanismus 2013. Austrian Energy Agency.

Bergsten, P., Nicolin, S., Frisk, A., Rydgren, B., Poveda Björklund, I., & Stage, J. (2014). Miljöförbättrande åtgärder i Mörrumsån och Ångermanälven, med fokus på havsvandrande arter. ÅF Industry AB.

Brooks, T., & Balmford, A. (1996). Atlantic forest extinctions. Nature, 380(6570), 115.

Chen, A., Wu, M., Chen, K. Q., Sun, Z. Y., Shen, C., & Wang, P. Y. (2016). Main issues in research and practice of environmental protection for water conservancy and hydropower projects in China. Water Science and Engineering, 9(4), 312–323.

Cheng, F., Li, W., Castello, L., Murphy, B. R., & Xie, S. (2015). Potential effects of dam cascade on fish: Lessons from the Yangtze River. Reviews in Fish Biology and Fisheries, 25(3), 569–585.

Cooper, A. R., Infante, D. M., Wehrly, K. E., Wang, L., & Brenden, T. O. (2016). Identifying indicators and quantifying large‐scale effects of dams on fishes. Ecological Indicators, 61, 646–657.

Diamond, J. M. (1972). Biogeographic kinetics: Estimation of relaxation times for avifaunas of Southwest Pacific islands. Proceedings of the National Academy of Sciences of the United States of America, 69(11), 3199–3203. Diamond, J. M., & May, R. M. (1981). Island biogeography and the design of nature reserves. In R. M. May (Ed.),

Theoretical ecology: Principles and applications(2nd ed., pp. 228–252). Oxford: Blackwell.

Drakare, S., Lennon, J. J., & Hillebrand, H. (2006). The imprint of the geographical, evolutionary and ecological context on species–area relationships. Ecology Letters, 9(2), 215–227.

Eadie, J. M., Hurly, T. A., Montgomerie, R. D., & Teather, K. L. (1986). Lakes and rivers as islands_ species‐area relationships in the fish faunas of Ontario. Environmental Biology of Fishes, 15(2), 81–89.

European Commission. (2015). Ecological flows in the implementation of the Water Framework Directive (Guidance Document 31). European Commission Directorate General for Environment.

Gan, L., Eskeland, G. S., & Kolshus, H. H. (2007). Green electricity market development: Lessons from Europe and the US. Energy Policy, 35(1), 144–155.

Hanski, I. (2011). Habitat loss, the dynamics of biodiversity, and a perspective on conservation. Ambio, 40(3), 248–255.

Harris, L. D. (1984). The fragmented forest: Island biogeography theory and the preservation of biotic diversity, Chicago: University of Chicago Press.

Hart, D. D., Johnson, T. E., Bushaw‐Newton, K. L., Horwitz, R. J., Bednarek, A. T., Charles, D. F., … Velinsky, D. J. (2002). Dam removal: Challenges and opportunities for ecological research and river restoration. BioScience, 52(8), 669–682.

Hasselquist, E. M., Nilsson, C., Hjältén, J., Jørgensen, D., Lind, L., & Polvi, L. E. (2015). Time for recovery of riparian plants in restored northern Swedish streams: A chronosequence study. Ecological Applications, 25(5), 1373–1389.

Hedenskog, M., & Monsén, J. (2012). Omprövning av vattendomar. Publ nr 2012:12. Karlstad: Länsstyrelsen Värmland.

Håkansson, C. (2009). Costs and benefits of improving wild salmon passage in a regulated river. Journal of Environmental Planning and Management, 53, 345–363.

Isbell, F., Reich, P. B., Tilman, D., Hobbie, S. E., Polasky, S., & Binder, S. (2013). Nutrient enrichment, biodiversity loss, and consequent declines in ecosystem productivity. Proceedings of the National Academy of Sciences of the United States of America, 110(29), 11911–11916.

Kaunda, C. S., Kimambo, C. Z., & Nielsen, T. K. (2012). Hydropower in the context of sustainable energy supply: A review of technologies and challenges. ISRN Renewable Energy, 2012, 730631.

Kummer, M. (2002). Energieminderproduktion bei Wasserkraftwerken aufgrund der Restwasserbestimmungen im Gewässerschutzgesetz. Bern: OFEV.

MacArthur, R., & Wilson, E. O. (1967). The theory of island biogeography, Princeton: Princeton University Press. Mitchell, M. G. E., Suarez‐Castro, A. F., Martinez‐Harms, M., Maron, M., McAlpine, C., Gaston, K. J., … Johansen, J. R. (2015). Reframing landscape fragmentation's effects on ecosystem services. Trends in Ecology & Evolution, 30(4), 190–198.

Mueller, M., Pander, J., & Geist, J. (2011). The effects of weirs on structural stream habitat and biological communities. Journal of Applied Ecology, 48, 1450–1461.

O'Connor, J. E., Duda, J. J., & Grant, G. E. (2015). 1000 dams down and counting. Science, 348(6234), 496–497. Paulrud, A., & Laitila, T. (2013). A cost–benefit analysis of restoring the Em river in Sweden: Valuation of

angling sites and visitation frequency. Applied Economics, 45, 2255–2266.

Pimm, S. L., Russel, G. J., Gittleman, J. L., & Brooks, T. M. (1995). The future of biodiversity. Science, 269, 347–350.

Pindyck, R. S. (1991). Irreversibility, uncertainty, and investment. Journal of Economic Literature, 29, 1110–1152. Powers, R. P., & Jetz, W. (2019). Global habitat loss and extinction risk of terrestrial vertebrates under future

land‐use‐change scenarios. Nature Climate Change, 9(4), 323–329.

Rosenzweig, M. L. (1995). Species diversity in space and time. Cambridge: Cambridge University Press. Stage, J. (2015). The value of the Swedish eel fishery. Marine Resource Economics, 30(1), 21–34.

Stage, J. (2018). Living in a bubble: Potential gains from flexible water management policies. Applied Economics Letters, 25(19), 1368–1372.

Stigler, H., Huber, C., Wulz, C., & Todem, C. (2005). Energiewirtschaftliche und ökonomische Bewertung potenzieller Auswirkungen der Umsetzung der EU‐Wasserrahmenrightlinie auf die Wasserkraft. Institut für Elektrizitätswirtschaft und Energieinnovaton der Technischen Universität Graz.

Wearn, O. R., Reuman, D. C., & Ewers, R. M. (2012). Extinction debt and windows of conservation opportunity in the Brazilian Amazon. Science, 337(6091), 228–232.

Wilcove, D. S., McLellan, C. H., & Dobson, A. P. (1986). Habitat fragmentation in the temperate zone. Conservation Biology, 6, 237–256.

WWF. (2004). Dams and freshwater ecosystems—Repairing the damage. WWF Dams Initiative. Retrieved from http://www.panda.org/dams

Xu, W., Ouyang, Z., Viña, A., Zheng, H., Liu, J., & Xiao, Y. (2006). Designing a conservation plan for protecting the habitat for giant pandas in the Qionglai mountain range, China. Diversity and Distributions, 12(5), 610–619.

Ziv, G., Baran, E., Nam, S., Rodríguez‐Iturbe, I., & Levin, S. A. (2012). Trading‐off fish biodiversity, food security, and hydropower in the Mekong River Basin. Proceedings of the National Academy of Sciences of the United States of America, 109(15), 5609–5614.

How to cite this article: Akpalu W, Stage J. Connectivity at a cost: Economic dynamics of restoring habitat connectivity. Natural Resource Modeling. 2020;e12294.