Transnational Ecosystem Services: The Potential of Habitat Conservation for Waterfowl Through Recreational Hunting Activities

Abstract

This article explores transnational ecosystem services in North America, provided by winter habitat for waterfowl in western Mexico coastal lagoons, and the hunting industry supported by these birds in the United States. This article shows that the number of waterfowl harvested in the United States is related to the abundance of waterfowl wintering in Mexico. On average, this flow of ecosystem services annually yields US$ 4.68 million in hunting stamp sales in the western United States. A demand curve, fitted to duck hunting licenses as a function of stamp price and previous-year waterfowl harvest, estimated US$3–6 million in consumer surplus produced in addition to governmental stamp sales revenue. This strongly suggests that waterfowl wintering habitat in western Mexico is economically valuable to U.S. hunters. Because hunters may benefit substantially from these habitats they may be willing to pay for conservation efforts in western Mexico that can result in transnational benefits received in the United States.

Keywords:

• transnational economic benefits
• wetlands
• Gulf of California
• coastal lagoons
• mangrove
• conservation

Introduction

Habitat deterioration worldwide is arguably one of the most critical drivers for the creation of initiatives that link ecological processes and economic benefits. It is widely recognized that migratory birds are vulnerable to habitat deterioration, since they depend on various distant habitats during their life cycle (Marra, Hobson, & Holmes, 1998; Sultanian & Van Beukering, 2008). Additionally, migratory birds require functional ecological attributes along continental latitudes, in order to fulfill their migratory needs and to accomplish their role as ecosystem service providers (Whelan, Wenny, & Marquis, 2008). Besides the recreational activities (e.g., hunting and bird watching) that these birds provide, which are the focus of this article, they generate a wide range of ecosystem services through their functional role as predators, seed dispersers, and ecosystem engineers (Millennium Ecosystem Assessment, 2005; Whelan, Wenny, & Marquis, 2008). In spite of the economic benefits that can be gained from preserving migratory bird habitats for recreational activities, the willingness-to-pay to support habitat conservation efforts by the stakeholders beyond their national territories has been little explored.

Millions of migratory waterfowl cruise along the Pacific coast of North America. The United States, Canada, and Mexico are visited by 27 waterfowl species arriving mainly through the Pacific and Central Flyways. These migratory waterfowl winter in several tropical and subtropical areas, including coastal areas in western Mexico. On average, the west coast of Mexico receives between 7 and 17% of all the migratory waterfowl in North America every year (North American Waterfowl Management Plan [NAWMP], 2004).

Waterfowl wintering in coastal lagoons depends strongly on the organic matter input from these wetlands (e.g., mangroves and sea grasses) and the habitat these lagoons provide (Altenburg & Spanje, 1989; Lefebvre & Poulin, 1996; Marra, Hobson, & Holmes, 1998). Wintering coastal lagoons supply necessary conditions for critical waterfowl life stages such as feeding, courtship, and molting (Bellrose & Low, 1978; Lovelock & Ellison, 2007; Saunders & Saunders, 1981; Wilson & Ryan, 1997). In spring, the waterfowl fly back to their northern breeding grounds and are a source of economic incentives for the recreational hunting and tourism industries in the United States and Canada (Carver, 2008).

Historically, the United States and Canada have been involved in conservation initiatives for migratory birds, as a part of international cooperation agreements between both countries since the beginning of the twentieth century. Additionally, a 1916 convention between the United States and Great Britain produced an agreement for the Protection of Migratory Birds (Fjetland, 2000). In 1932, the United States created the Migratory Bird Conservation Commission and the Conservation Fund. The latter obtains funding from the revenues of sales of federal duck stamps, which were created in 1934 as required licenses for hunting migratory waterfowl. Ninety-eight cents of every dollar goes directly to buy or lease wetland habitat in the United States (Bolen, 2000; U.S. Fish and Wildlife Service, n.d.). Through this system, federal duck stamps have so far contributed to the purchase and protection of more than 21,000 km² of wetlands and habitat for the U.S. National Wildlife Refuge System (National Wildlife Refuge Association, n.d.).

Waterfowl conservation actions in Mexico, however, did not begin until several decades later, in 1993, when Mexico joined the North American Free Trade Agreement (NAFTA) (Fjetland, 2000; NAWMP, 2004). Mexico is now also a signatory party to the North American Waterfowl Management Plan (NAWMP), which has been successful in the United States and Canada through a joint venture (JV) system, or collaboration to achieve NAWMP goals at a regional scale. Through JV efforts, more than 44,500 km² have been protected and recovered in the United States and Canada (NAWMP, 2007). However, Mexico is not a part of the JV system or other organizational structures critical to effectively preserve wetland habitat and recover waterfowl populations. Wetlands in Mexico are disappearing at a national rate of 7% annually because of sedimentation, eutrophication, and deforestation (National Commission for Knowledge and Use of Biodiversity [CONABIO], 2009), and are facing increasing pressures to transform these habitats into shrimp farms and tourism developments (Glenn, Nagler, Brusca, & Hinojosa-Huerta, 2006; Páez-Osuna et al., 2003).

After the droughts of the 1980s and 1990s in the U.S.–Canada Prairie Pothole Region ended, populations of the majority of duck species important for hunting activities are exhibiting rising trends (Wilkins, Otto, & Smith, 2000). This increase in waterfowl populations has sustained a profitable U.S. recreational hunting industry in the last decades. Even though the number of hunters in the United States has slightly declined in the last few decades, revenue generated from duck hunting licenses, or duck stamps, remains a critical economic resource for the conservation of waterfowl populations and habitat in the United States (Carver, 2008; NAWMP, 2007). For example, in 2011 there were 13.7 million people in the United States engaged in hunting activities, from which 2.6 million were migratory waterfowl hunters (U.S. Fish and Wildlife Service [USFWS], 2012). The economic benefits of duck-stamp purchases represent only a fraction of other associated expenditures, including trips to locations, hunting equipment, job incomes, and state and federal tax revenues, all of which together generate billions of dollars annually (Carver, 2008). The estimated average expenditure of a hunter in 2011, represented US$2,484, an increase of 24% over the previous decade (USFWS, 2012).

This article tests the hypothesis that there is a positive correlation between the winter abundance of migratory waterfowl in Mexico, the harvest rates of these birds in the United States and the associated duck stamp sales. The economic implication of this hypothesized relationship is that a flow of benefits is sent from the wintering habitats to consumers of hunting and recreational activities in the United States. The existence of this cross-border ecosystem service in North America suggests there might be losses of benefits if habitat size or quality decline.

Methods

Long-Term Monitoring Surveys Data and Conceptual Framework

Cooperative efforts between the United States, Canada, and Mexico have consolidated five major waterfowl monitoring programs with annual surveys, which together comprise the largest dataset of any wildlife group in the world (Blohm, Sharp, Padding, & Richkus, 2006). Since the 1950s, the USFWS has consolidated five major waterfowl monitoring programs with annual surveys on populations, production (breeding), habitat, banding, and harvest (Blohm et al., 2006).

Wintering waterfowl abundance and habitat in the Gulf of California for this study were described by coupling two datasets. Firstly, a database obtained from USFWS that includes Mexican midwinter waterfowl aerial surveys from 1947 through 2006 along the Pacific coast, Central Mexico, and Gulf of Mexico, was analyzed by digitizing the airplane track lines (USFWS, 2010). The digitized tracks denote nine study sites surveyed for waterfowl counts wintering in the Gulf of California Figure 1). Four sites are single lagoons or small bays: (a) Bahía de Santa Bárbara, (b) Estero de Lobos, (c) Laguna Huizache-Caimanero, and (d) Laguna Agiabampo. The other sites are lagoon complexes: (e) Tiburón Island to Guaymas, (f) Guaymas to Estero de Lobos, (g) Bahía de San Esteban and Topolobampo, (h) Bahía de Santa María, Ensenada de Pabellones, and El Dorado to Dimas, and (i) Marismas Nacionales and Laguna Cuyutlán Table 1).

Figure 1. Gulf of California coastal lagoons (black circles) included in the Pacific Flyway. Gray color represents mangrove cover and dotted lines represent the aircraft track line where migratory wildfowl were counted (see Methods). Coastal lagoons or lagoon complexes are: (a) Tiburón Island to Guaymas, (b) Guaymas to Estero de Lobos, (c) Estero de Lobos, (d) Bahía de Santa Bárbara, (e) Laguna Agiabampo, (f) Bahía de San Esteban/Topolobampo, (g) Bahía de Santa Maria/Ensenada de Pabellón/El Dorado to Dimas, (h) Laguna Huizache-Caimanero, and (i) Marismas Nacionales/Laguna Cuyutlán (color figure available online).

The Mexican mid-winter aerial surveys were done annually from 1947 until 1985, and every 3 years thereafter. The longest, almost uninterrupted, sets of data were collected from 1977 to 1996. This time period of 19 years was used to analyze the average abundance of 17 waterfowl species present in the database. Since five species (Northern Pintail, Northern Shoveler, American Green-winged Teal, Blue-winged/Cinnamon Teal, and American Wigeon) represent 85% of total abundance, the analysis presented here focuses on the time series trend of these species (Table 1). The TRIM software (Trends and Indices for Monitoring data, version 3.04) was used to calculate annual population estimates for the time series of the aerial count data of waterfowl wintering in the Gulf of California, following the methodology described by Pannekoek and VanStrain (2005).

Table 1 Study site information and estimated average abundance of the five species of waterfowl analyzed in this study

				Estimated average abundance (1977–1996)**
Site name	Map ID*	Lagoon size (km^2)	Mangrove fringe (km^2)	NOPI	NOSHO	AGWT	BWCT	AWIG	Total
Tiburon Island to Guaymas	a	170.08	13.29	1,472.30	0	0	0	0	1,472.30
Guaymas to Estero de Lobos	b	97.4	56.61	1,030.50	0	227.4	0	0	1,257.80
Estero de Lobos	c	137.4	8	1,953.40	1,204.60	2,621.90	282.8	925.8	6,988.50
Bahia de Santa Barbara	d	78.12	7.03	9,055.20	7,550.40	10,452.00	2,072.70	4,927.10	34,057.30
Laguna Agiabampo	e	184.79	12.63	16,466.10	6,581.90	12,641.40	1,996.60	5,634.10	43,320.00
Bahia de San Esteban to Topolobampo	f	678.46	221.17	53,152.00	66,552.90	62,198.80	28,225.20	25,886.30	236,015.10
Bahia de Santa Maria, Ensenada de Pabellon, and El Dorado to Dimas	g	1,070.52	420.64	185,538.30	70,927.50	42,716.60	58,333.30	11,501.90	369,017.40
Laguna Huizache Caimanero	h	153.61	10.88	13,632.60	15,421.30	3,121.50	2,104.50	4,346.50	38,626.30
Marismas Nacionales to Laguna Cuyutlan	i	285.8	825.55	27,520.70	56,648.10	16,584.60	21,997.90	3,777.50	126,528.80
Notes. *ID (mangrove study site identification letter) as shown in map Figure 1. **Estimated average abundance (1977–1996); species labels: Northern Pintail (NOPI), Northern Shoveler (NOSHO), American Green-winged Teal (AGWT), Bluewinged/Cinnamon Teal (BWCT), American Wigeon (AWIG).

Second, in order to digitize the extent of the areas surveyed during the winter monitoring program the aerial track lines followed during the waterfowl counts were analyzed (Figure 1; Mallek, Wortham, & Eldridge, 2008). Lagoon size and wetland cover (particularly mangrove habitat) within these digitized polygons were estimated using published spatial data from Landsat Thematic Mapper (TM) satellite images from years 2003 to 2007, covering 60 km × 60 km of terrestrial surface, and with a minimum pixel size of 10 meters (Carrera & de la Fuente, 2003; CONABIO, 2008).

Waterfowl harvest data were obtained with a USFWS database initiated in 1952. The USFWS has conducted a daily harvest survey of Federal Duck Stamp buyers, to estimate migratory bird hunter activity and harvest at the state, flyway, and national levels. The Cooperative State-Federal Migratory Bird Harvest Information Program (HIP) generates reliable estimates of national harvest levels of migratory birds in the United States. States ask hunters screening questions about the species they hunt and their hunting success the previous year. Starting in July and through the end of the hunting season, the states send information about migratory bird hunters to the USFWS. After the initial batch of names and addresses is received from a state, the USFWS selects a random sample of hunters in each state, and mails them a survey form. Thousands of survey forms are sent across the country. Hunters are asked to record the date of each hunt, location, and how many ducks and geese they hunted for each day of the hunting season. In the year 2009, for the Pacific Flyway, 8,230 surveys were sent to hunters and the average response rate was 56% (B. Raftovich, personal communication, March 24, 2010). The USFWS uses the hunters’ responses on the survey forms to estimate harvest and hunter activity (Raftovich, Wilkins, Richkus, Williams, & Spriggs, 2009). For the Pacific Flyway, harvest trends of 14 species are available at the Flyway U.S. website (Flyways.us, n.d.).

Estimating Direct Revenue from Hunting Stamps

To study economic outcomes, a Duck Stamp Sales and Revenue database from the USFWS Division of Migratory Bird Management was used (USFWS, unpublished data provided by B. Raftovich, personal communication, March 24, 2010; see Table 2). This database contains the cost of the stamp and the number of Federal Duck Stamps sold each year from 1934 through 2008. Data were adjusted to 2010 U.S. dollars using the U.S. Bureau of Labor Statistics inflation calculator (USBLS; Table 2). Duck stamp sales were used from the states belonging to the U.S. Pacific flyway, which include: Arizona, California, Idaho, Nevada, Oregon, Utah, and Washington.

Table 2 Historical abundance and harvest numbers of waterfowl, with the historical sales and prices of duck stamp within the Pacific flyway of North America

			Stamps
Year	Wintering waterfowl (1977–96) ^§	Waterfowl harvested ^†	Sold	Historical price US$ ^‡	2010 Price US$ ^¥
1977	795,688	1,724,910	352,279	5.0	17.95
1978	1,495,290	2,263,930	350,179	5.0	16.70
1979	1,255,760	1,918,780	338,113	7.5	22.50
1980	990,110	1,675,340	332,736	7.5	19.80
1981	746,720	1,279,570	309,215	7.5	17.93
1982	1,118,460	1,462,030	312,706	7.5	16.95
1983	830,223	1,418,120	311,110	7.5	16.43
1984	992,440	1,084,290	298,623	7.5	15.68
1985	868,554	1,038,230	273,889	7.5	15.15
1986	645,177	993,940	248,846	7.5	14.93
1987	560,502	1,056,510	220,205	10.0	19.20
1988	540,760	574,140	186,226	10.0	18.74
1989	645,317	706,440	186,682	12.5	22.00
1990	804,296	676,340	179,031	12.5	20.88
1991	1,045,865	646,870	171,051	12.5	20.00
1992	884,250	648,160	169,435	12.5	19.38
1993	812,351	769,630	193,424	12.5	18.88
1994	803,240	834,010	191,242	15.0	22.05
1995	625,520	1,199,420	205,745	15.0	21.45
1996	524,699	1,355,750	205,970	15.0	20.85
^§Abundance (1977–1996) column shows the sum of the estimated abundances for each of the five species per year in the nine study sites.
^†Harvest data values represent the sum of the harvest values in the United States per year for each of the five species in this study.
^‡Historical Price US$ shows the price paid for each duck stamp at a given year.
^¥2010 US$ prices are the transformed historical prices in 2010 US$ values. Data were converted using the U.S. Bureau of Labor Statistics inflation calculator.

Using standard time-series analysis techniques, trends in the inflation-adjusted price of hunting stamps were tested. First, the time series was tested for the existence of a generalized, upward or downward, linear trend. Afterward, the main trend residuals of the price series were tested for the existence of a periodic cycle in adjusted stamp values. For this purpose, we used both an autocorrelation analysis and Fourier analysis (Chatfield, 2003).

A simple conceptual framework was followed to explore the association between all variables through time using Pearson correlations. Relationships between variables were fitted using reduced major axes regression, as in all cases both the independent and the dependent variables were subject to sampling error. The importance of lagoon size as waterfowl habitat was explored by correlating the area of each of the nine lagoon systems and the extent of fringe mangrove against the cumulative, 20-year average abundance of the five most common waterfowl species described previously. Because lagoon area and mangrove extent are positively correlated, path analysis (described below) was employed to detect the relative importance of each variable.

All other variables analyzed corresponded to yearly data collected between 1977 and 1996, and were analyzed with a Path Analysis (Everitt & Dunn, 1991). These variables included, for each year: (a) the number of waterfowl counted in the nine lagoons, (b) the number of birds harvested by hunters, (c) the number of federal duck stamps sold, and (d) the price of the stamps. Path Analysis, originally developed by Sewal Wright (1921), is a combination of a graphic representation of the main functional pathways between variables and simple algebraic manipulations based on multiple regression that allow testing for alternative relational hypotheses on the system under study (for a detailed explanation on the theory of path analysis see Everitt & Dunn, 1991; and Sokal & Rohlf, 1995). Taking stamp sales as the final variable to be predicted, the path coefficients of stamp sales were calculated against all others, and retained only those predictors whose path coefficients proved to be significant. Finally, the best predictors of the selected predictors of stamp sales were searched and, using the same method, the most significant predictors were found for each variable moving up the path. Path coefficients (β) are really partial correlation coefficients of each predictor on the predicted variable; as such, they range between zero and unity and are always less than, or equal to, the direct correlation coefficient (r) between pairs of variables.

This analysis also tested whether there was a time lag between the number of birds harvested against the number of stamps sold. Namely, two hypotheses were tested, (a) that the number of birds harvested in a given season is correlated to the number of stamps sold during that same season and (b) that the number of birds harvested in any given year may influence future expectations and hence is correlated to the number of stamps sold during the following hunting season. For this purpose, path analysis was performed, as described previously, on the original 20-year data series, and then repeated lagging stamp sales and price a year behind waterfowl wintering abundance counts and waterfowl harvest numbers.

Valuing Economic Surplus

Finding a path relationship between the ecological health of the lagoons and stamp sales revenue is a good first approach at quantifying the value of the environmental services provided by wintering wetlands for migratory waterfowl as an economic resource. However, the revenue generated by stamp sales is only a part of the economic services provided by these wetlands for waterfowl hunting as an economic activity. Stamp sales are collected by the USFWS and, as such, they measure income for the U.S. government but do not necessarily reflect the true value of ecosystem services provided to the hunters themselves, which may be higher. A rough estimate of the uncollected value of these services can be obtained by calculating the consumer surplus (also known as the Marshallian surplus; Marshall, 1920) of the activity, which is the monetary gain obtained by consumers when they are able to purchase a product (the right to hunt) for a cost (the stamp price) that is less than the price that they would be willing to pay.

In order to estimate the willingness-to-pay price we developed a demand curve by regressing stamp sales, as dependent variable, against stamp price and previous-year harvest, as independent predictors. Using this model, we then projected the maximum price buyers would be theoretically willingness-to-pay by solving for the price values that make the fitted stamp sales reach zero. We calculated the willingness-to-pay prices of hunting stamps (i.e., the price intercept) under three scenarios: (a) a minimum harvest during the previous year (0.57 million ducks), (b) an average previous-year harvest (1.16 million ducks), and (c) a maximum harvest (2.26 million). Once we had obtained the estimates of the willingness-to-pay price for stamps, we calculated the area under the demand curve, which is simply equal to the area of a triangle under the line: CS = q.(p_max – p_mean )/2, where CS is the consumer's surplus value, q is the mean number of stamps sold every year, p_mean is the mean value of each stamp, and p_max is the willingness-to-pay price estimated from the demand curve (Mankiw, 2011).

We also fitted, through multiple regression, a demand curve on the log-transformed variables, by regressing log-stamp sales against the logarithms of both stamp price and previous-year harvest. The coefficients of the log-transformed regression are elasticity values, and can be interpreted as the ratio between the relative rate of change of the dependent variable against each independent variable (Binger & Hoffman, 1997). If a given coefficient is less than unity, the dependent variable changes less than the independent predictor when the latter is modified. Coefficient values higher than unity indicate that the dependent variable is highly sensitive to changes in the independent predictor, and will vary more, in relative terms, than the independent predictor when the predictor changes. Thus, these log-regression coefficients are direct measures of the economic elasticity of stamp sales against stamp price and hunting expectations (the previous-year harvest), and allow the testing of hypotheses on the relative importance of factors having an incidence on stamp sales.

Results and Discussion

Abundance of Migratory Waterfowl, Harvest Rates, and Stamp Sales

The water-body area of the nine lagoon sites was significantly correlated with waterfowl abundance (r = .98, p < .0001; Figure 2a), while, in absolute terms, the amount of mangrove fringe was not (r = .57, p = .09). However, the path analysis showed that, after fixing the effect of lagoon area as a covariate, the amount of mangrove fringe was also significantly related to waterfowl numbers (p = .03 for a t-test on the path coefficient; see Figure 3). That is, the simple lagoon area is the most important driver of waterfowl abundance, but the amount of mangrove fringe—an indicator of the health of the wetland—increases the lagoon's capacity to provide habitat for migratory birds. On average, each square kilometer of coastal lagoon harbors some 348 counted birds using the wetland as wintering habitat; and, given a certain lagoon size, each additional square kilometer of mangrove fringe increases the survey counts by 65 birds (birds = 348 × lagoon area + 65 × fringe – 26,684; F _3,6 = 140.5; p < .0001).

Figure 3. Path analysis: The width of the black arrows and the associated numbers indicate the values of the significant path coefficients. The dotted ellipses correspond to the path analysis of lagoon size and mangrove extent as predictors of waterfowl counts. All other ellipses (continuous lines) correspond to the variables analyzed and significant paths identified as predictors of stamp sales. Total revenue was simply calculated as the product of stamp sales × inflation-adjusted stamp prices. White arrows show path points where other, external variables may have an effect.

Figure 2. Correlation between the variables analyzed in this study: (a) Relationship between wintering waterfowl abundance and coastal lagoon area in ten Mexican lagoons (r = .98; n = 9 lagoons); (b) annual waterfowl harvest in the United States plotted against migratory waterfowl abundance in Mexico (r = .65; n = 19 years); (c) hunting stamps sold yearly against previous-year waterfowl harvest in the United States (r = .90; n = 19 years); (d) annual stamp-sale revenue against hunting stamps sold in the United States (r = .86; n = 19 years). All correlations were significant at the P = .01 level.

The time-lagged model between the abundance of waterfowl counted in their wintering lagoons and the number of birds harvested provided better correlation values and path coefficients against the number of stamps sold during the following year than the un-lagged model, supporting the second hypothesis that the number of birds harvested in a year influences future expectations and is correlated to the number of stamps sold during the following hunting season.

All yearly variables analyzed (wintering waterfowl counts and birds harvested by hunters, and stamps sold and stamp price during the following year) were significantly correlated between themselves, with the exception of stamp price that was only negatively correlated to the number of stamps sold (r = −.44; n = 19 years; p = .05). After these correlations were simplified using path analysis (Figure 3), the significant path coefficients that remained showed: (a) that the amount of stamps sold was positively associated with the number of birds harvested during the previous year and negatively associated with stamp price and (b) that the number of harvested birds was highly associated with waterfowl counts during the preceding winter.

Transnational Economic Benefits: Stamp-Sale Revenue

Under standard economic assumptions, the principal determinants of stamp purchases by hunters should be stamp price and expectations of hunting opportunities. In agreement with this hypothesis, lagged coefficients between previous year harvests (a reasonable proxy for hunters’ expectations) and stamp sales were highly significant predictors of stamp sales. For the past 30 years, waterfowl hunting activities in the western United States have contributed to consolidate an industry in which, on average, US$4.68 million annually are generated through duck stamp sales (in 2010 inflation-adjusted values). Our results show that, on average, an extra wintering bird counted in the previous year was associated with an increase in duck stamp sales by US$4.75 (y = 4.75 x + .48 × 10⁶; p < .001, r ² = .67). Using the slope of the linear model of lagoon area against wintering duck counts (y = 380.7x – 25.5 × 10³; p < .001, r ² = .98; Figure 2a), we can estimate that a decrease of 1 km² of coastal lagoon in Mexico may lead to a loss of some US$1,640 in duck hunting stamp revenues in the US (se = ± 140 US$, calculated using a jackknife estimate).

Transnational Economic Benefits: Consumer Surplus

A multiple regression model using the previous year's harvest as a proxy for the expected harvest in the current year and stamp price as the second predictor, showed a highly significant fit of the demand curve to the stamp sales data (r ² = .90; F _3,16 = 69.5; p < .00001). Expectedly, we found that willingness to pay (i.e., the price intercept) increases with the previous year's harvest: If previous year's harvest is good, then the demand curve predicts that hunters will be more willing to pay for higher stamp prices during the current year (Table 3).

Table 3 Willingness-to-pay (WTP) price estimates for stamp sales calculated under three scenarios: (i) minimum, (ii) average, and (iii) maximum previous-year harvest. The consumer surplus was calculated from the product of the WTP price and the mean number of stamps sold every year (see Methods for details)

	Previous year harvest
	Minimum	Average	Maximum
Harvested birds	574,140	1,156,351	2,263,930
Willing-to-pay price (US$)	42.25	50.97	67.56
Consumer surplus (US$)	2,878,395	3,953,323	5,998,235

Knowing that the average annual duck stamps sold is ca. 252,000, the estimated Marshallian surplus varies between US$3 and 6 million according to the hunters’ expectations of a successful season, which, in turn, seem to be mostly driven by the abundance of waterfowl during the previous year. Even though it cannot be claimed this surplus is fully extractable, however, it represents a rough estimate of excess willingness-to-pay for stamps that may serve to shape expectations for the scope of payments for ecosystem services (PES) at a transnational scale (Lopez-Hoffman, Varady, Balvanera, & Flessa, 2010).

Elasticity and Sensitivity of Stamp Sales

The elasticity of stamp sales with respect to stamp price was low (.44 ± .18), and differed significantly from the expected value of unity (p < .009). The elasticity of stamp sales with respect to previous-season harvest was somewhat higher, but still significantly lower than unity (.54 ± .06; p < .0001). These results show that stamp sales are fairly inelastic and tend to maintain their numbers even when their prices vary. They also confirm the results of the path analysis, in the sense that the previous year waterfowl harvest has a larger bearing on stamp sales than year-to-year variations in stamp price.

On the other hand, price variations in stamp sales seem to be low and driven by exogenous factors, as the time-series analysis showed that the inflation-adjusted prices of stamps has no significant long-term trend and no significant periodicity. In fact, the adjusted data shows that stamp prices were kept roughly constant by government for at least 20 years (Figure 4).

Figure 4. Nominal and inflation-adjusted price values (US$) for duck stamps between 1977 and 1996. The horizontal dotted line marks the time-series average.

Finally, as described in the Methods section, the willingness-to-pay price of hunting stamps was estimated by fitting a demand function to our data set, and then extrapolating to the point where sales become zero. Like with any other extrapolation, this value is calculated outside the realm of the observed data and is hence subject to errors. Bearing this in mind, it is sufficient for our purposes to conclude that the economic surplus not collected in the form of stamp revenue may be as high as, or perhaps higher than, the value of the stamp sales.

Conclusion

The results of this study suggest that the hunting industry in the United States responds to the abundance of the waterfowl that migrate to Mexico, which in turn depend on the size and health of the lagoon habitat in their wintering grounds. This means that one square kilometer of coastal lagoon in Mexico generates US$1,640 in duck hunting stamp revenues in the United States. Furthermore, these waterfowl hunting activities will have an influence on the number of stamps sales in the following hunting season, since U.S. hunters appear to receive benefits from hunting in excess of their hunting stamp expenditures. Besides the economic revenues generated by stamp sales in the United States on the order of US$4 million, there are economic benefits for waterfowl hunters which had an estimated Marshallian surplus between US$3 and 6 million. These surplus benefits suggest a possible source of funding to conserve habitat beyond the borders of the United States.

Additionally, this study contributes to ongoing research on quantifying economic benefits of ecosystems. While data may not always exist to put precise “price tags” on habitats on a per-area basis, it remains important to record and disseminate the available information. A final take-home message of this study is that an economic industry in the United States is supported by an ecosystem in Mexico, through nearly one million birds that winter in the eastern side of the Gulf of California. The willingness-to-pay for stamps by U.S. hunters certainly exceeds that which they do pay, and may encourage plans for establish a Payment of Ecosystem Services (PES) at a transnational scale. These private funds could motivate Mexican government agencies to combine efforts with the United States and Canada for the establishment of legislative, logistic, and economic mechanisms needed to establish a robust and widespread collaborative structure through Mexico's territory.

As was mentioned in the introduction, the JV system has been successful in achieving regional-scale conservation goals for migratory birds in the United States and Canada, but Mexico has lagged behind in the establishment of this system. A functional transnational PES, strongly supported by the U.S. hunters’ willingness-to-pay for habitat conservation, could be an opportunity to expedite the creation of a Mexican JVs as an important strategy to preserve biodiversity and increase conservation efforts in Mexico (Salcido, Quiroz, & Ramírez, 2009). In the Baja California Peninsula, for instance, land is already being purchased through the support of the North American Wetlands Conservation Act (NAWCA) in collaboration with local nongovernmental organizations (NGOs). Through bird-watching ecotourism, as well, the private sector is generating regional jobs and promoting transnational habitat conservation (Carver, 2009; Connell, 2009); an example that could inspire the hunting sector to preserve habitat beyond U.S. borders.

However, the implementation of a PES to establish a JVs structure along Mexico's territory will represent an important challenge, since the country has a centralized government and most of the times the conservation decisions are taken with a top-down approach, and tend to ignore the local social realities happening at the regional scale. The most successful cases of PES are in the developed world, where the idea of a “Coasean approach” in which property rights are well defined (Muradian et al., 2010) has allowed the establishment and success of these programs. Contrary, in developing countries, the implementation of PES program will have to take a more dynamic approach, because the difference related with the social structures, the lack of enforce in property rights, and the minimal institutional support for these programs (Clements et al., 2010).

Although the idea of transnational PES seems to be an unrealistic endeavor given the actual governmental scheme, Mexico already has received economic support from the World Bank to establish two main PES programs: (1) The Program of Payment for Hydrological Environmental Services of Forests (PSAH), which consists of direct payments to landowners with forests in good state of conservation and (2) The Program for the Development of Markets for the Ecosystem Services of Carbon Sequestration, the Derivatives of Biodiversity, and to Promote the Introduction and Improvement of Agro-forestry Systems (PSA-CABSA). These PES programs are currently some of the world's largest and most ambitious in that respect, since multiple ecosystem components and services are involved: water, biodiversity, carbon sequestration, and agro-forestry services (McAfee & Shapiro, 2010). We believe it is possible that these kinds of PES projects can be tailored for coastal ecosystems and wetlands, and presented to the government agencies with a similar structural frame as the ongoing PES projects. And given the precedent set by voluntary contributions to groups such as Ducks Unlimited performing conservation work in Mexican wetlands (Ducks Unlimited, n.d.), hunters on the United States and even Canada may be willing to contribute to create this large, long-term PES for coastal lagoons and wetlands in Mexico, after viewing these results.

Acknowledgments

We thank M. Otto, B. Raftovich, and K. Richkus for their assistance with the databases and literature. I. Fogel of the Centro de Investigaciones Biológicas del Noroeste provided valuable editorial comments.

Funding

Funding for this study was provided to O.A.O. and C.H.G.A. by the David and Lucile Packard Foundation and the Walton Family Foundation. N.T.R.C. is the recipient of a national graduate fellowship from the Consejo Nacional de Ciencia y Tecnología of Mexico (CONACYT) and the UC Institute for Mexico and the United States (UC MEXUS).

Notes

© Nadia T. Rubio-Cisneros, Octavio Aburto-Oropeza, Jason Murray, Charlotte E. Gonzalez-Abraham, Jeremy Jackson, Exequiel Ezcurra.