At Quitobaquito Springs in Organ Pipe Cactus National Monument, a desert pond smaller than a city block holds the only wild populations of the Sonoyta pupfish, the Quitobaquito springsnail, and the Sonoyta mud turtle. There is nowhere else on Earth where these three species live together in the wild. The Department of Homeland Security has awarded a contract to Sundt Construction to build a second 30-foot border wall 90 to 150 feet from the springs, and according to U.S. Customs and Border Protection, work could commence at any moment.

During construction of the first wall in 2020, the pond partially dried up after its lining cracked, and the spring’s flow fell to an all-time low. That episode — machinery operated near the pond, the lining failed, water escaped, the flow dropped — is causation with a directly observed physical mechanism, not a statistical inference. The second wall, placed even closer to the water, threatens to repeat and compound the same failure mode.

The structural cause is neither a single administration nor a particular contractor. It is Section 102(c) of the Illegal Immigration Reform and Immigrant Responsibility Act, as amended by the Real ID Act, which grants the DHS secretary unilateral authority to waive “all legal requirements” the secretary determines necessary to ensure expeditious construction of border barriers. Without this waiver, the project near Quitobaquito Springs would have required an environmental impact statement under the National Environmental Policy Act, formal Section 7 consultation with the U.S. Fish and Wildlife Service under the Endangered Species Act, and tribal consultation under the Native American Graves Protection and Repatriation Act. The waiver removes the procedural safeguards that would force a documented assessment of extinction risk before ground is broken. The siting decision — building a second wall 90 to 150 feet from the pond — is a permitted outcome of a legal framework that makes such siting legally costless.

The aquifer that feeds the springs is supplied by extremely old groundwater and water from the Rio Sonoyta to the south — a system with negligible natural replenishment in a hyper-arid environment. Contractors use large volumes of groundwater to mix cement for the wall’s base. CBP estimated that a single mile of fence built nearby in 2008 consumed about 710,000 gallons of water; current projects likely require more because the new walls are larger. Heavy machinery for the second wall will operate 90 to 150 feet from the pond, whose southern edge sits 190 feet from the international border. A grove of mesquite trees closer to the border, part of federally designated critical habitat for the mud turtle, is likely to be leveled. Michael Bogan, an aquatic ecologist at the University of Arizona who leads the Sonoyta mud turtle recovery team, warned that heavy groundwater use for construction could lower the water table below the point where water exits the spring, causing it to run dry. “That concern keeps me up,” he said. Tara Harris, director of conservation and science at the Phoenix Zoo, put it directly: “It could destroy the last remaining habitat for multiple imperiled species.”

The absence of site-specific hydrological data is itself a product of the waiver — NEPA review would have required baseline water-table measurements, impact modeling, and cessation triggers. The drawdown rate from Sundt’s cement operations is unknown; the aquifer’s recharge profile is inferred from regional hydrogeology rather than measured at the site. CBP spokesperson John Mennell said the agency was “working with the US Fish and Wildlife Service and National Park Service to avoid Quitobaquito Springs and minimize impacts to sensitive resources.” CBP declined to answer follow-up questions.

The power structure around this decision is stark. DHS and Sundt Construction hold the authority and the contract; the three endemic species hold the highest biological stake and zero agency. The National Park Service and U.S. Fish and Wildlife Service have a consultation role only, with no independent authority to halt construction under the waiver. The University of Arizona recovery team has domain expertise but no halt authority. The O’odham people have relied on the springs for generations; the waiver bypasses NAGPRA, leaving them without the tribal consultation that statute would normally require. Wynona Larson Yazzie, an O’odham elder and community organizer, called the planned construction “sickening.” No tribunal, court order, or inter-agency arbitration mechanism sits between the waiver authority and the ground.

The backup-colony strategy is a hedge without assurance of return. The Arizona-Sonoran Desert Museum holds 20 Sonoyta mud turtles and has agreed to accept up to 50 more, per herpetology curator Tom Weaver. The Phoenix Zoo maintains the only backup colony of Quitobaquito springsnails, housed in two tanks, and would also take in turtles. Combined museum-and-zoo capacity covers at most 70 turtles — fewer than 30 percent of the estimated 250 wild turtles at the spring. Two snail tanks cannot replicate the spring’s specific water chemistry. Debra Colodner, the Desert Museum’s director of conservation science, stated the core constraint: “Our main concern is that there’s a place for them to go back to.” These are stopgaps, not a viable insurance population for species with no other wild habitat. Mexico is an implied stakeholder because the aquifer is supplied by water from the Rio Sonoyta, a binational watercourse. If the aquifer is depleted, communities drawing from the Rio Sonoyta system in Sonora could also be affected — a binational consequence with no existing coordination mechanism documented in the record.

The waiver’s reach extends well beyond this single spring. According to CBP, 34 miles of secondary barriers have been completed, mostly in Arizona, creating two adjacent 30-foot walls. An additional 390 miles of barriers are under construction and 155 miles have been awarded contracts. Erick Meza, borderlands coordinator with the Sierra Club, described similar secondary fencing at about six other locations on the Arizona-Sonora border. “I’ve never seen something of this magnitude. It’s unbelievable,” he said. The waiver is the operating procedure, not a one-time exception, for a border-barrier program that has become structurally exempt from environmental law.

Two independent uncertainties determine what happens next. One is legal: whether the waiver regime survives. The other is hydrological: whether the aquifer physically survives the pumping. Together they produce four possible trajectories.

Hardened Perimeter. The waiver remains intact, and the aquifer sustains the pumping. Construction proceeds at the planned distance. Groundwater pumping draws the water table down but not below the outflow point. The spring’s flow declines but stabilizes at a reduced rate; the pond shrinks but does not dry. Backup colonies are maintained but do not become permanent. The species persist in the wild at diminished numbers, and the legal-exemption precedent holds for the six other sites.

Mitigation Corridor. The waiver is curtailed — by a legal challenge, an administration change, or congressional action — before construction reaches the springs, and the aquifer proves resilient. The project is forced through environmental review, halted or redesigned with groundwater monitoring and setback requirements. The 2020 damage proves recoverable. Bogan’s team shifts from evacuation planning to habitat restoration. O’odham tribes gain a formal consultation role.

Losing Both. The waiver remains intact, and the spring fails. Construction proceeds. Groundwater pumping lowers the water table below the outflow point. The spring runs dry. The Sonoyta pupfish and Quitobaquito springsnail become extinct in the wild — no other habitat on Earth exists for either. The Sonoyta mud turtle survives only in captivity. The mesquite grove, federally designated critical habitat, is leveled. The O’odham lose one of their sacred springs. The extinction is foreseeable and is not a legal violation because the waiver is in effect — the statute grants the permission, the result is biological erasure.

Legal Reversal. The waiver is curtailed, but the spring fails anyway. Pumping has already lowered the water table below the outflow point before curtailment arrives. The pupfish and springsnail are gone. The mud turtle persists only in captivity. The precedent forces re-evaluation at the six other secondary-barrier sites. The Real ID Act’s waiver provision faces a repeal effort. The spring may or may not recover over decades if pumping stops.

There is a wild card that sits outside all four quadrants. If the same hydrogeological risk — heavy groundwater pumping for cement drawing down a fossil aquifer — materializes at two or more of the six other Arizona-Sonora border sites simultaneously, the result is cascading loss of multiple desert wetlands, multiple endemic species, and multiple sacred O’odham sites at once. The backup-colony system is designed for single-site failure; a multi-site collapse would overwhelm it. Quitobaquito is a test case, not an isolated incident.

Three framing questions will answer themselves over the coming months. Can a statutory waiver survive a legal challenge grounded in irreversible species loss? Is the aquifer’s natural replenishment rate high enough to absorb construction pumping without crossing the threshold where the spring stops flowing? And if the spring fails at Quitobaquito, does the same mechanism threaten the six other sites where identical secondary fencing is being built under identical legal authority? The indicators that will answer these questions are hydrological, not political. Spring flow data from the University of Arizona and the National Park Service will show whether the aquifer is absorbing or failing under construction demand. A decline exceeding 20 percent from pre-construction baselines during active work would signal the worst-case scenario is materializing and that the window for evacuating the mud turtles is closing. A federal court injunction citing Section 102(c) waivers as overbroad would signal the most favorable outcome. A CBP announcement pausing the Sundt Construction contract would be another favorable signal. CBP’s pattern of declining to answer follow-up questions about hydrological impact tracks the worst-case trajectory — the institution responsible for measuring the damage declines to measure it.

Regardless of which scenario unfolds, several actions are urgent and robust across all four. Backup colony capacity at the Desert Museum and Phoenix Zoo must be expanded beyond current commitments — 20 turtles and two springsnail tanks are not a viable insurance population for species with no wild alternative. Continuous groundwater and spring-flow monitoring must be installed before construction begins, not after. Baseline species counts, water chemistry, and piezometer levels must be documented now for any future legal or restoration work. If spring flow monitoring shows a decline exceeding 20 percent from pre-construction baselines during active construction, Sonoyta mud turtles should be evacuated to the Desert Museum without waiting for political resolution — the window closes fast in the worst-case scenario.

The corrective fix is narrow: require site-specific hydrological assessment and construction proximity limits at Quitobaquito Springs before work proceeds, through DHS administrative action or congressional restriction. The preventive fix is structural: amend the Real ID Act to preserve environmental law compliance within designated critical habitat and national monument boundaries, restoring the review processes the waiver bypasses for ecologically irreplaceable sites. The waiver is the operating procedure, not a one-time exception. Until that changes, Quitobaquito is not the last spring at risk — it is the first one with a name we know.

Analytical techniques used in this piece

This analysis applies the methods below. Each links to a short, plain-English explainer you can read and reuse.

Root-Cause Analysis
Traces a symptom back along its causal chain to the conditions that actually generated it.
Scenario Planning
Builds a small set of distinct, plausible futures to plan against.
Stakeholder Mapping
Charts the parties to a situation — their interests, power, and alignments.