Effectiveness of Parks in Protecting Tropical Biodiversity

See allHide authors and affiliations

Science  05 Jan 2001:
Vol. 291, Issue 5501, pp. 125-128
DOI: 10.1126/science.291.5501.125


We assessed the impacts of anthropogenic threats on 93 protected areas in 22 tropical countries to test the hypothesis that parks are an effective means to protect tropical biodiversity. We found that the majority of parks are successful at stopping land clearing, and to a lesser degree effective at mitigating logging, hunting, fire, and grazing. Park effectiveness correlates with basic management activities such as enforcement, boundary demarcation, and direct compensation to local communities, suggesting that even modest increases in funding would directly increase the ability of parks to protect tropical biodiversity.

Much of tropical biodiversity is unlikely to survive without effective protection (1–3). Conservationists have tried strategies ranging from establishing and maintaining parks and other strictly protected areas (henceforth “parks”), to promoting sustainable forest management and other integrated conservation and development projects. How well do parks measure up among these alternatives (4, 5)? Critics claim that in the context of growing human pressures and development needs, parks cannot protect the biological resources within their borders (6), and there is a widespread sense that parks are simply not working. The accuracy of these claims is of critical importance to policy and funding decisions. If parks are failing despite best efforts, then better options should be sought. If, on the other hand, parks are performing relatively well in a context of serious threats and limited resources, or are simply performing better than the alternatives, their level of support should be increased.

Past studies of park effectiveness have focused on improving park management (7), improving protected area system design (8), and increasing local and national political support (9, 10), but none has provided a quantitative assessment of effectiveness using a large sample of parks around the world (11, 12).

We used a questionnaire to collect data on land-use pressure (land clearing, logging, hunting, grazing, and fire), local conditions (e.g., presence of human communities in parks and degree of access), and management activities (e.g., number of guards and level of community involvement in management) (13). To confine our sample to parks at risk of failure, we selected regions subject to significant human land-use pressure (14, 15). From these regions, we selected only parks that have been established for at least 5 years to allow sufficient time for management activities to be reflected in park performance. We also restricted the sample to protected areas of >5000 ha in which only nonconsumptive uses were permitted (16, 17). Directors of conservation organizations and protected area agencies helped identify a representative group for this study from the 535 parks that met these criteria (18, 19) [additional information is available on Science Online (20)]. The sample comprised 93 parks (21) in 22 countries (22), covering 17% (18 million ha) of the parks that met our criteria (23).

The parks in the sample varied greatly in size, primary ecosystem type, budget, management strategy, and type and degree of threats. Seventy percent had people living inside their boundaries, and 54% had residents who contested the ownership of some percentage of the park area. Two-thirds of the parks were accessible by at least one major road or river (24). Median annual funding was 1.18 USD ha−1, significantly less than the amount often recommended for effective management (25). Finally, respondents judged that many park staff were lacking in critical training and equipment.

We assessed the effectiveness of these parks from three perspectives: land clearing within the boundaries of parks since establishment, current condition of parks compared with the condition of their surroundings, and factors correlated with effective park protection.

We assessed the effectiveness of parks at preventing land clearing by comparing the current extent of clearing with clearing at the time of park establishment (Fig. 1). We found that 43% of the parks have had no net clearing since establishment. In an additional 40% of parks, land formerly under cultivation was incorporated into park boundaries, and had been able to recover, leading to an actual increase in vegetative cover. Eighty-three percent of parks were therefore fully holding their borders against agricultural encroachment. Only 17% of the parks experienced net clearing since establishment. This is a substantial achievement, given that the median age of the parks in our sample is 23 years.

Figure 1

Change in the area of natural vegetation since establishment for 86 tropical parks. The majority of parks have either experienced no net clearing or have actually increased natural vegetative cover. Median park age is 23 years.

To test effectiveness over a wider range of threats, we compared anthropogenic impacts in the 10-km belt surrounding parks with the level of impacts within park boundaries for five different threats (Fig. 2). This comparison shows that the parks in our sample are under great pressure from clearing, hunting, and logging, and to a lesser extent, fire and grazing. A comparison of the conditions inside the parks with the surrounding area shows that for all five threats, parks were in significantly better condition than their surrounding areas (Mann-Whitney U-test, medians significantly different atP < 0.006 for all five impacts). Because we used relatively few response categories to represent the entire range of outcomes (e.g., four categories were used to classify the abundance of game animals, ranging from pristine levels of abundance to absent), any differences found between the parks and their surroundings are great.

Figure 2

Comparison of the condition of parks to the surrounding 10-km belt. For all five anthropogenic impacts, parks were in significantly better condition than their surrounding areas (Mann-Whitney U-test) (A) Clearing: P = 0.000; (B) logging: P = 0.000; (C) hunting: P = 0.000; (D) fire:P = 0.000; and (E) grazing:P = 0.006.

Parks are more effective at mitigating some impacts than others. Parks are in far better condition than their surroundings with respect to land clearing, with the majority of parks being intact or only slightly cleared. Parks were more heavily impacted by logging and hunting, but these impacts were still reduced considerably compared with their surroundings. Finally, although parks were still in significantly better condition than their surroundings with respect to damage from fire and grazing, the differences were less pronounced.

The previous comparison treats the sample of parks as a group. We also compared individual parks with their own surroundings to determine what percentage of individual parks are functioning (Table 1). Virtually all parks in our sample are under pressure from clearing, grazing, fire, hunting, and logging, and the majority of parks are effective at least to some degree in mitigating these threats. More than 80% of the individual parks were in better condition than their surroundings for clearing, logging, and fire, including 97% for clearing. About 60% of the parks were in better condition than land outside their borders with respect to hunting and grazing.

Table 1

Park effectiveness against anthropogenic threats. Shown for each threat is the percentage of parks surveyed that are either in better condition (“functioning”) than the surrounding 10-km belt or in equal or worse condition (“not functioning”). Also shown is the percentage of parks with no presence of each threat in the surrounding area (“untested parks”).

View this table:

We also investigated which management activities and local conditions correlated with effectiveness, which we defined as the difference between illegal impacts inside the park and the surrounding 10-km belt. Because units and scales differed among the threats, we rescaled each threat to a value ranging from 1 to 4, and then averaged the effectiveness among the five threats. Average effectiveness was then tested for correlation with management activities and local conditions by means of a nonparametric test (26).

Park effectiveness correlated most strongly with density of guards (Table 2). The median density of guards in the 15 most effective parks was more than eight times higher than in the 15 least effective parks (3 guards per 100 km2 in the 15 most effective parks compared with 0.4 guards per 100 km2 in the least effective). However, enforcement capacity (a composite variable of training, equipment, and salary) was not found to correlate with effectiveness, suggesting that these characteristics are less important than the presence of guards.

Table 2

Spearman rank correlations of variables with park effectiveness. Variables that are significantly correlated with effectiveness are shown in bold in the upper portion of the table. Where sample size and type of data permitted, the ratio of the 15 most effective to 15 least effective parks is included to illustrate the difference in magnitude.

View this table:

Effectiveness was also significantly correlated with the level of deterrents to illegal activities in the park. Deterrents were measured as the product of the probability of apprehending violators when guards detected a violation (either in progress or after-the-fact) by the probability of the violator receiving a significant sanction if apprehended. Deterrents against clearing and logging correlated with park effectiveness, whereas deterrents against hunting did not.

The degree of border demarcation and the existence of direct compensation programs to local communities (Table 2) were also found to correlate significantly with management effectiveness.

Other factors potentially related to park success did not correlate significantly with effectiveness, including number of people living in the park, accessibility, local support, percentage of the park area contested, budget, number of staff working on economic development or education, and local involvement of communities in park management.

The findings of this study suggest three basic conclusions. First, the claim that the majority of parks in tropical countries are “paper parks”—i.e., parks in name only—is not substantiated. Tropical parks have been surprisingly effective at protecting the ecosystems and species within their borders in the context of chronic underfunding and significant land-use pressure. They have been especially effective in preventing land clearing, arguably the most serious threat to biodiversity. Second, despite their successes, there is a clear need to increase support for parks to improve effectiveness against all threats, perhaps especially against hunting. Finally, these findings suggest that parks should remain a central component of conservation strategies. Both creating new parks and addressing the tractable problem of making existing parks perform better will make a significant contribution to long-term biodiversity conservation in the tropics.

  • * To whom correspondence should be addressed. E-mail: a.bruner{at}


View Abstract

Navigate This Article